Field notes on machine recommendation
How to read the result
One answer to the role-only question is one ballot. Geography splits the races: a Czech ranking and a worldwide ranking are different elections and never share a tally. Nothing here says who is best — only who AI products name when a stranger asks.
The role alone, authored in several wordings, crossed with search enabled and disabled.
Electorate: person-agnostic recommend prompts only
Each answer's names, in the order the answer put them, with a numbered slot told apart from a prose mention.
Watch: honourable-mention tails
Two spellings of one person split their own vote. Near-duplicates are reported, never merged.
Watch: shared surnames, shared profiles
Borda where the stage has run; otherwise a mention count, labelled as one.
Watch: variants that disagree
The read
Reading the ballots…
The electorate is the person-agnostic recommend family; rung-3/4 and recognition answers are excluded from every tally on this page.
The headline
One race per geography. The bar is the leader's score on whichever tally this run has — Borda where the stage has run, ballots naming them otherwise.
Answers contributed · names returned per answer · search arms
01 / Search arms
The two arms side by side, per election, over the products that ran both ways. This measures the effect of enabling search: a search-on model is handed a capability, not compelled to use it, so this is never memory against live retrieval.
02 / Disparity
Two products can answer the same question with one name in common. This panel measures that distance and does not judge it: a name only one product offers may be a name only one product knows. Every share is printed beside the number of ballots it was measured over.
03 / Ballots
How long the answers ran against what was asked for, whether a name sat in a numbered slot or in prose, and how many Subject Runs were kept out of the electorate. This is the panel the pilot gate is read from — next to the archive.
04 / Names
Near-duplicate names, the aliases the tally declined to merge, and every name the answers put forward. Under a ranking a split clone does not merely under-count one person — it can hand the top slot to someone else.
05 / Sources
Which domains each product consulted while building its ballots, and how much searching it did. Run grain: no page here is attributed to any candidate.
06 / Spend
Subject-call volume, errors, token load, and modeled API cost. Grading and analysis calls remain outside this total.
05 / Run archive
Each shared prompt is shown once. Open it to compare products, search modes, and samples; then expand any response for Markdown, grading, citations, cost, and the raw record.
| borda | dowdall | length_normalized | n_ballots | n_products | best_rank | unresolved_alias_flag | |
|---|---|---|---|---|---|---|---|
| person | |||||||
| Jana Bártíková | 1.347 | 0.234 | 0.089 | 25 | 3 | 1 | False |
| Michaela Chramostová | 0.917 | 0.147 | 0.061 | 12 | 4 | 1 | False |
| Michelle Losekoot | 0.875 | 0.151 | 0.058 | 12 | 3 | 1 | False |
| Martin Hošek | 0.639 | 0.104 | 0.041 | 9 | 4 | 1 | False |
| Eliška Vyhnánková | 0.472 | 0.059 | 0.031 | 13 | 4 | 2 | False |
| Sergej Pavljuk | 0.444 | 0.056 | 0.028 | 10 | 3 | 2 | False |
| Radka Legerská | 0.417 | 0.058 | 0.027 | 9 | 3 | 1 | False |
| Katka Lusková | 0.389 | 0.059 | 0.026 | 7 | 3 | 1 | False |
| Filip Dřímalka | 0.278 | 0.036 | 0.019 | 4 | 2 | 2 | False |
| Miloš Křepelka | 0.278 | 0.056 | 0.019 | 3 | 2 | 1 | False |
| Adam Zbiejczuk | 0.264 | 0.037 | 0.018 | 8 | 4 | 2 | False |
| Vojtěch Forejtek | 0.250 | 0.042 | 0.017 | 2 | 1 | 1 | False |
| Lenka Kazdová | 0.250 | 0.043 | 0.017 | 4 | 1 | 1 | False |
| Jiří Jambor | 0.236 | 0.028 | 0.015 | 6 | 3 | 2 | True |
| Klára Zbončáková | 0.222 | 0.027 | 0.015 | 4 | 2 | 3 | False |
| Daniel Gamrot | 0.222 | 0.029 | 0.015 | 3 | 1 | 2 | False |
| Josef Kadlec | 0.208 | 0.025 | 0.014 | 8 | 2 | 2 | True |
| Jan Cina | 0.208 | 0.042 | 0.014 | 1 | 1 | 1 | False |
| Jan Řezáč | 0.194 | 0.035 | 0.013 | 2 | 1 | 1 | False |
| Jan Tegze | 0.194 | 0.023 | 0.013 | 3 | 2 | 2 | False |
| Barbora Křížová | 0.153 | 0.022 | 0.010 | 4 | 2 | 2 | False |
| Marie Zámečníková | 0.139 | 0.028 | 0.009 | 1 | 1 | 1 | False |
| Michal Šrajer | 0.139 | 0.028 | 0.009 | 1 | 1 | 1 | False |
| Hana Kavanová | 0.139 | 0.016 | 0.009 | 3 | 1 | 3 | False |
| David Kolář | 0.139 | 0.016 | 0.009 | 2 | 1 | 3 | False |
| Jan Tůma | 0.125 | 0.014 | 0.008 | 1 | 1 | 3 | False |
| Petra Drahoňovská | 0.125 | 0.016 | 0.008 | 3 | 2 | 3 | False |
| Jan Kyselý | 0.125 | 0.014 | 0.007 | 3 | 2 | 3 | False |
| Lucie Spáčilová | 0.111 | 0.014 | 0.007 | 1 | 1 | 2 | False |
| Radek Váňa | 0.111 | 0.014 | 0.007 | 1 | 1 | 2 | False |
| Radim Hasalík | 0.111 | 0.014 | 0.007 | 1 | 1 | 2 | False |
| Martin Schwarz | 0.097 | 0.013 | 0.006 | 3 | 2 | 2 | False |
| Petr Casanova | 0.083 | 0.010 | 0.006 | 1 | 1 | 4 | False |
| Tomáš Jindříšek | 0.083 | 0.009 | 0.006 | 1 | 1 | 3 | False |
| Barbora Křivová | 0.083 | 0.009 | 0.006 | 1 | 1 | 3 | False |
| Pavel Šíma | 0.083 | 0.009 | 0.006 | 1 | 1 | 3 | False |
| Marek Prchal | 0.083 | 0.012 | 0.006 | 2 | 1 | 4 | False |
| Jan Bartoš | 0.069 | 0.014 | 0.005 | 1 | 1 | 1 | False |
| Lucie Zitterbartová | 0.069 | 0.014 | 0.005 | 1 | 1 | 1 | False |
| Kateřina Štěpánková | 0.056 | 0.007 | 0.004 | 1 | 1 | 4 | False |
| Igor Buinevici | 0.056 | 0.007 | 0.004 | 1 | 1 | 4 | False |
| Jan Řežáb | 0.056 | 0.007 | 0.004 | 1 | 1 | 4 | False |
| Ivo Toman | 0.056 | 0.007 | 0.004 | 1 | 1 | 4 | False |
| Jarmila Vindušková | 0.056 | 0.007 | 0.004 | 2 | 1 | 4 | False |
| Eva Mátlová | 0.056 | 0.007 | 0.004 | 1 | 1 | 4 | False |
| Ilona Smelík | 0.056 | 0.007 | 0.004 | 2 | 1 | 4 | False |
| Boris Vaňka | 0.056 | 0.007 | 0.004 | 1 | 1 | 4 | False |
| Pavlína Louženská | 0.042 | 0.008 | 0.003 | 2 | 2 | 5 | False |
| Tereza Kubíčková | 0.042 | 0.005 | 0.003 | 2 | 2 | 3 | False |
| Matěj Matolín | 0.042 | 0.008 | 0.003 | 1 | 1 | 5 | False |
| Tereza (LinkedIn konzultace a personal branding) | 0.042 | 0.005 | 0.003 | 1 | 1 | 3 | False |
| Alžběta Fridrichová | 0.028 | 0.006 | 0.002 | 2 | 2 | 5 | False |
| Helena Hájková | 0.028 | 0.006 | 0.002 | 1 | 1 | 5 | False |
| Lenka Kosmatová | 0.028 | 0.006 | 0.002 | 1 | 1 | 5 | False |
| Marie Krátká | 0.028 | 0.006 | 0.002 | 1 | 1 | 5 | False |
| Martin Kadlec | 0.028 | 0.006 | 0.002 | 1 | 1 | 5 | True |
| Pavel Cahlík | 0.028 | 0.006 | 0.002 | 1 | 1 | 5 | False |
| Petr Mareš | 0.028 | 0.003 | 0.002 | 1 | 1 | 4 | False |
| Petr Mára | 0.028 | 0.006 | 0.002 | 1 | 1 | 5 | False |
| Šárka Risch | 0.028 | 0.006 | 0.002 | 1 | 1 | 5 | False |
| Kateřina Žáková | 0.028 | 0.003 | 0.002 | 1 | 1 | 4 | False |
| Petr Hovorka | 0.014 | 0.003 | 0.001 | 1 | 1 | 5 | False |
| Petr Schwank | 0.014 | 0.003 | 0.001 | 1 | 1 | 5 | False |
| Lucie Václavková | 0.014 | 0.003 | 0.001 | 1 | 1 | 5 | False |
| Lukáš Vyskočil (Laskavý chachar) | 0.014 | 0.003 | 0.001 | 1 | 1 | 5 | False |
| Lenka Rožnovská | 0.014 | 0.003 | 0.001 | 5 | 1 | 5 | False |
| Kristýna Szkanderová | 0.000 | 0.000 | 0.000 | 2 | 1 | <NA> | False |
| Jiří Jambor (a agentura Future Sales) | 0.000 | 0.000 | 0.000 | 1 | 1 | <NA> | True |
| Lucie Königová | 0.000 | 0.000 | 0.000 | 1 | 1 | <NA> | False |
| Martina Špatná | 0.000 | 0.000 | 0.000 | 1 | 1 | <NA> | False |
| Michal Plichta | 0.000 | 0.000 | 0.000 | 1 | 1 | <NA> | True |
| Ondřej Merta | 0.000 | 0.000 | 0.000 | 1 | 1 | <NA> | False |
| Ondřej Souček | 0.000 | 0.000 | 0.000 | 1 | 1 | <NA> | False |
| Pavla Petrů | 0.000 | 0.000 | 0.000 | 1 | 1 | <NA> | False |
| Polina Ševčíková | 0.000 | 0.000 | 0.000 | 1 | 1 | <NA> | False |
| Tomáš Vachuda | 0.000 | 0.000 | 0.000 | 1 | 1 | <NA> | False |
| Tomáš Černovský | 0.000 | 0.000 | 0.000 | 1 | 1 | <NA> | False |
| Vendula Beránková | 0.000 | 0.000 | 0.000 | 1 | 1 | <NA> | False |
| Vojtěch Strnad | 0.000 | 0.000 | 0.000 | 1 | 1 | <NA> | False |
| Václav Šulista | 0.000 | 0.000 | 0.000 | 1 | 1 | <NA> | False |
Borda tally from borda.csv. Where the variants (Borda, Dowdall, length-normalized) disagree, that disagreement is the finding — not a reason to pick one._ _Pooled from linkedin-expert rung 1 (54 ballots). One election is one geography, which is how borda.csv keys its own tally — but a rung-2 ask is narrower than a rung-1 one and two intents ask two questions, so more than one family here means this list answers more than one question. Scope the run, or read it as pooled.
| borda | dowdall | length_normalized | n_ballots | n_products | best_rank | unresolved_alias_flag | |
|---|---|---|---|---|---|---|---|
| person | |||||||
| Marta Pálešová | 2.167 | 0.377 | 0.140 | 21 | 4 | 1 | False |
| Ivana Brutenič | 1.483 | 0.247 | 0.091 | 19 | 3 | 1 | True |
| Miki Plichta | 1.017 | 0.182 | 0.063 | 17 | 3 | 1 | True |
| Kristína Cichý Kováčiková | 0.933 | 0.133 | 0.057 | 13 | 4 | 1 | False |
| Michal Pastier | 0.550 | 0.084 | 0.037 | 10 | 2 | 1 | False |
| Oliver Jakubík | 0.550 | 0.067 | 0.038 | 8 | 4 | 2 | False |
| Lukáš Bednár | 0.483 | 0.063 | 0.030 | 9 | 4 | 2 | False |
| Michal Tramita | 0.433 | 0.054 | 0.029 | 5 | 2 | 2 | False |
| Daniel Výcha | 0.400 | 0.051 | 0.025 | 5 | 1 | 2 | False |
| Peter Chodelka | 0.300 | 0.050 | 0.020 | 4 | 1 | 1 | False |
| Lucia Mokráňová | 0.300 | 0.036 | 0.020 | 3 | 1 | 2 | False |
| Lucia Pekaríková | 0.283 | 0.035 | 0.019 | 7 | 2 | 3 | False |
| Peter Šebo | 0.283 | 0.037 | 0.019 | 6 | 2 | 2 | False |
| Andrea Vadkerti | 0.267 | 0.046 | 0.018 | 3 | 2 | 1 | False |
| Filip Kuna | 0.233 | 0.033 | 0.016 | 4 | 2 | 1 | False |
| Martin Volek | 0.217 | 0.036 | 0.014 | 4 | 3 | 1 | False |
| Dávid Macej | 0.200 | 0.026 | 0.013 | 3 | 1 | 3 | False |
| Ján Stareček | 0.200 | 0.022 | 0.013 | 2 | 1 | 3 | False |
| Mário Biháry | 0.167 | 0.033 | 0.011 | 1 | 1 | 1 | False |
| Peter Štefančík | 0.167 | 0.033 | 0.011 | 1 | 1 | 1 | False |
| Peter Kolesár | 0.167 | 0.033 | 0.011 | 1 | 1 | 1 | False |
| Zuzana Ďuricová | 0.167 | 0.023 | 0.012 | 2 | 1 | 2 | False |
| Lucia Mokrá | 0.167 | 0.033 | 0.011 | 1 | 1 | 1 | False |
| Kristína Kubíková | 0.167 | 0.033 | 0.008 | 1 | 1 | 1 | False |
| Martin Višňovský | 0.133 | 0.017 | 0.009 | 1 | 1 | 2 | False |
| Noema Pšenicová | 0.133 | 0.017 | 0.009 | 1 | 1 | 2 | False |
| Peter Kmoško | 0.133 | 0.017 | 0.009 | 1 | 1 | 2 | False |
| Peter Zvirinský | 0.133 | 0.017 | 0.009 | 1 | 1 | 2 | False |
| Tomáš Vrtík | 0.133 | 0.017 | 0.009 | 1 | 1 | 2 | False |
| Adam Stečák | 0.100 | 0.015 | 0.008 | 2 | 1 | 4 | False |
| Marek Šulík | 0.100 | 0.011 | 0.007 | 2 | 1 | 3 | False |
| Ján Záborský | 0.100 | 0.011 | 0.007 | 1 | 1 | 3 | False |
| Lucia Klestincová | 0.100 | 0.011 | 0.007 | 1 | 1 | 3 | False |
| Martin Krupa | 0.100 | 0.011 | 0.007 | 1 | 1 | 3 | False |
| Miroslav Reiter | 0.100 | 0.011 | 0.007 | 1 | 1 | 3 | False |
| Peter Krajčovič | 0.100 | 0.011 | 0.007 | 1 | 1 | 3 | False |
| Peter Strcula | 0.083 | 0.017 | 0.006 | 1 | 1 | 1 | False |
| Michal Kopecký | 0.083 | 0.017 | 0.006 | 1 | 1 | 1 | False |
| Sergej Pavljuk | 0.083 | 0.010 | 0.005 | 3 | 1 | 3 | False |
| Tony Dúbravec | 0.083 | 0.010 | 0.006 | 4 | 1 | 3 | False |
| Michal Holub | 0.067 | 0.008 | 0.004 | 1 | 1 | 4 | False |
| Zuzana Čmelíková | 0.067 | 0.008 | 0.004 | 1 | 1 | 4 | False |
| Taťána Mráziková | 0.067 | 0.008 | 0.004 | 1 | 1 | 4 | False |
| Peter Mészáros | 0.067 | 0.008 | 0.004 | 1 | 1 | 4 | False |
| Peter Murár | 0.067 | 0.008 | 0.004 | 1 | 1 | 2 | False |
| Tomáš Török | 0.067 | 0.008 | 0.004 | 1 | 1 | 2 | False |
| Jaroslav Dodok | 0.067 | 0.008 | 0.004 | 1 | 1 | 2 | False |
| Ivana Brutenič a Oliver Brutenič | 0.067 | 0.008 | 0.004 | 1 | 1 | 2 | True |
| Pavol Škorpík | 0.067 | 0.008 | 0.004 | 2 | 2 | 4 | False |
| Monika Filipová | 0.067 | 0.013 | 0.004 | 2 | 1 | 5 | False |
| Kristína Cichý Kováčová | 0.067 | 0.009 | 0.004 | 4 | 1 | 3 | False |
| Juraj Sasko | 0.050 | 0.010 | 0.003 | 3 | 1 | 5 | False |
| Tomáš Kromka | 0.050 | 0.010 | 0.003 | 3 | 1 | 5 | False |
| Katarína Košarišťanová | 0.050 | 0.008 | 0.003 | 2 | 1 | 4 | False |
| Gabriel Tóth | 0.050 | 0.006 | 0.003 | 1 | 1 | 3 | False |
| Kika Nevedelová | 0.050 | 0.006 | 0.003 | 1 | 1 | 3 | False |
| Mário Bihári | 0.033 | 0.007 | 0.002 | 1 | 1 | 5 | False |
| Tomáš Fülöp | 0.033 | 0.007 | 0.002 | 1 | 1 | 5 | False |
| Tomáš Bujna | 0.033 | 0.007 | 0.002 | 1 | 1 | 5 | False |
| Roland Vojkovský | 0.033 | 0.004 | 0.002 | 1 | 1 | 4 | False |
| Nina Krovinová | 0.033 | 0.004 | 0.002 | 1 | 1 | 4 | False |
| Anna Heveryová | 0.033 | 0.007 | 0.002 | 1 | 1 | 5 | False |
| Michal Gula | 0.033 | 0.004 | 0.002 | 1 | 1 | 4 | False |
| Marek Zeman | 0.033 | 0.007 | 0.002 | 1 | 1 | 5 | False |
| Karin Zvalová | 0.033 | 0.007 | 0.002 | 1 | 1 | 5 | False |
| Juraj Zaťko | 0.033 | 0.007 | 0.002 | 1 | 1 | 5 | False |
| Andrej Kajan | 0.033 | 0.007 | 0.002 | 1 | 1 | 5 | False |
| Hana Jech | 0.017 | 0.003 | 0.001 | 1 | 1 | 5 | False |
| Peter Sámel | 0.017 | 0.003 | 0.001 | 1 | 1 | 5 | False |
| Sergej Danilov | 0.017 | 0.003 | 0.001 | 1 | 1 | 5 | False |
| Simona Mištíková | 0.017 | 0.003 | 0.001 | 1 | 1 | 5 | False |
| Veronika Hesková | 0.017 | 0.003 | 0.001 | 1 | 1 | 5 | False |
| Kristína Cichý Brutenič | 0.000 | 0.000 | 0.000 | 1 | 1 | <NA> | True |
| Martin Hoško | 0.000 | 0.000 | 0.000 | 1 | 1 | <NA> | False |
| Michal Ukropec | 0.000 | 0.000 | 0.000 | 1 | 1 | <NA> | False |
| Slavomír Bača | 0.000 | 0.000 | 0.000 | 1 | 1 | <NA> | False |
| Ľubomíra Mardiaková | 0.000 | 0.000 | 0.000 | 1 | 1 | <NA> | False |
Borda tally from borda.csv. Where the variants (Borda, Dowdall, length-normalized) disagree, that disagreement is the finding — not a reason to pick one._ _Pooled from linkedin-expert rung 1 (54 ballots). One election is one geography, which is how borda.csv keys its own tally — but a rung-2 ask is narrower than a rung-1 one and two intents ask two questions, so more than one family here means this list answers more than one question. Scope the run, or read it as pooled.
| borda | dowdall | length_normalized | n_ballots | n_products | best_rank | unresolved_alias_flag | |
|---|---|---|---|---|---|---|---|
| person | |||||||
| Richard van der Blom | 3.111 | 0.526 | 0.207 | 24 | 6 | 1 | False |
| Justin Welsh | 3.061 | 0.548 | 0.204 | 23 | 6 | 1 | False |
| Jasmin Alić | 1.917 | 0.241 | 0.128 | 18 | 5 | 1 | False |
| Lara Acosta | 1.444 | 0.184 | 0.096 | 16 | 5 | 2 | False |
| Goldie Chan | 0.444 | 0.064 | 0.030 | 10 | 4 | 4 | False |
| Dorie Clark | 0.356 | 0.047 | 0.024 | 4 | 2 | 2 | False |
| Michaela Alexis | 0.333 | 0.046 | 0.022 | 6 | 3 | 2 | False |
| Andre Santos | 0.333 | 0.051 | 0.022 | 3 | 2 | 1 | False |
| Melonie Dodaro | 0.300 | 0.035 | 0.020 | 3 | 2 | 3 | False |
| Viveka von Rosen | 0.278 | 0.039 | 0.019 | 5 | 2 | 2 | False |
| Andy Foote | 0.222 | 0.030 | 0.015 | 3 | 1 | 3 | False |
| Amelia Sordell | 0.200 | 0.028 | 0.013 | 5 | 4 | 2 | True |
| John Espirian | 0.167 | 0.019 | 0.011 | 2 | 2 | 3 | False |
| William Arruda | 0.133 | 0.018 | 0.009 | 2 | 1 | 3 | False |
| Daniel Disney | 0.111 | 0.018 | 0.007 | 5 | 2 | 4 | False |
| Bhavik Sarkhedi | 0.111 | 0.022 | 0.007 | 2 | 1 | 5 | False |
| Fredrik Fornes | 0.111 | 0.014 | 0.007 | 1 | 1 | 2 | False |
| Michelle J Raymond | 0.083 | 0.009 | 0.006 | 1 | 1 | 3 | False |
| Sam G. Winsbury | 0.083 | 0.009 | 0.006 | 1 | 1 | 3 | False |
| Lea Turner | 0.083 | 0.009 | 0.006 | 1 | 1 | 3 | False |
| Joshua B. Lee | 0.083 | 0.012 | 0.006 | 3 | 2 | 4 | False |
| Matt Barker | 0.056 | 0.011 | 0.004 | 2 | 1 | 5 | False |
| Blossom Affia | 0.056 | 0.007 | 0.004 | 1 | 1 | 4 | False |
| Judi Fox | 0.033 | 0.007 | 0.002 | 4 | 2 | 5 | False |
| Louise Brogan | 0.028 | 0.006 | 0.002 | 1 | 1 | 5 | False |
| Salina Yeung | 0.028 | 0.006 | 0.002 | 1 | 1 | 5 | False |
| Hanna Larsson | 0.000 | 0.000 | 0.000 | 1 | 1 | <NA> | False |
| Sam Browne (SocialSam) | 0.000 | 0.000 | 0.000 | 1 | 1 | <NA> | True |
| Sam Browne | 0.000 | 0.000 | 0.000 | 1 | 1 | <NA> | True |
| Molly Stovold | 0.000 | 0.000 | 0.000 | 1 | 1 | <NA> | False |
| Mandi McEwen | 0.000 | 0.000 | 0.000 | 1 | 1 | <NA> | False |
| Josh Fechter | 0.000 | 0.000 | 0.000 | 1 | 1 | <NA> | False |
| John Nemo | 0.000 | 0.000 | 0.000 | 1 | 1 | <NA> | False |
| Brynne Tillman | 0.000 | 0.000 | 0.000 | 1 | 1 | <NA> | False |
| Donna Serdula | 0.000 | 0.000 | 0.000 | 1 | 1 | <NA> | False |
| Carol Kaemmerer | 0.000 | 0.000 | 0.000 | 1 | 1 | <NA> | False |
| Ashraf Malik | 0.000 | 0.000 | 0.000 | 1 | 1 | <NA> | False |
| Arman Harutyunyan | 0.000 | 0.000 | 0.000 | 1 | 1 | <NA> | False |
| Amelia Sordell (Klowt) | 0.000 | 0.000 | 0.000 | 1 | 1 | <NA> | True |
| AJ Wilcox | 0.000 | 0.000 | 0.000 | 2 | 2 | <NA> | False |
| Tim Queen | 0.000 | 0.000 | 0.000 | 1 | 1 | <NA> | False |
Borda tally from borda.csv. Where the variants (Borda, Dowdall, length-normalized) disagree, that disagreement is the finding — not a reason to pick one._ _Pooled from linkedin-expert rung 1 (27 ballots). One election is one geography, which is how borda.csv keys its own tally — but a rung-2 ask is narrower than a rung-1 one and two intents ask two questions, so more than one family here means this list answers more than one question. Scope the run, or read it as pooled.
| ballots | search enabled | search disabled | mean_names | named_people | generic_advice | refusal | clarifying_question | |
|---|---|---|---|---|---|---|---|---|
| subject | ||||||||
| claude-sonnet | 30 | 15 | 15 | 2.97 | 17 | 13 | 0 | 0 |
| gemini | 30 | 15 | 15 | 5.27 | 30 | 0 | 0 | 0 |
| openai | 30 | 15 | 15 | 5.33 | 30 | 0 | 0 | 0 |
| deepseek | 15 | 0 | 15 | 6.00 | 15 | 0 | 0 | 0 |
| kimi | 15 | 0 | 15 | 2.13 | 7 | 7 | 1 | 0 |
| perplexity | 15 | 15 | 0 | 5.53 | 15 | 0 | 0 | 0 |
mean_names is what to read against what the treatment asked for. The two arms are search enabled and disabled: a search-on model is handed a WebSearch capability, never compelled to use it, so this is not memory versus live.
| enabled place | enabled ballots | disabled place | disabled ballots | place shift | |
|---|---|---|---|---|---|
| name | |||||
| Jana Bártíková | 1 | 14 | 1 | 11 | 0 |
| Radka Legerská | 2 | 9 | <NA> | 0 | <NA> |
| Martin Hošek | 3 | 6 | 7 | 2 | 4 |
| Sergej Pavljuk | 4 | 5 | 5 | 5 | 1 |
| Míša Chramostová | 4 | 5 | <NA> | 0 | <NA> |
| Jiří Jambor | 6 | 4 | 7 | 2 | 1 |
| Miloš Křepelka | 7 | 3 | <NA> | 0 | <NA> |
| Martin Schwarz | 8 | 3 | <NA> | 0 | <NA> |
| Adam Zbiejczuk | 9 | 3 | <NA> | 0 | <NA> |
| Jan Kyselý | 9 | 3 | <NA> | 0 | <NA> |
| Katka Lusková | 9 | 3 | <NA> | 0 | <NA> |
| Barbora Křížová | 12 | 3 | <NA> | 0 | <NA> |
| Vojtěch Forejtek | 13 | 2 | <NA> | 0 | <NA> |
| David Kolář | 14 | 2 | <NA> | 0 | <NA> |
| Eliška Vyhnánková | 15 | 2 | 2 | 8 | -13 |
| Alžběta Fridrichová | 16 | 2 | <NA> | 0 | <NA> |
| Lucie Zitterbartová | 17 | 1 | <NA> | 0 | <NA> |
| Michaela (Míša) Chramostová | 17 | 1 | <NA> | 0 | <NA> |
| Jan Tegze | 19 | 1 | 9 | 2 | -10 |
| Radek Váňa | 19 | 1 | <NA> | 0 | <NA> |
| Tereza (LinkedIn konzultace a personal branding) | 21 | 1 | <NA> | 0 | <NA> |
| Tereza Kubíčková | 21 | 1 | <NA> | 0 | <NA> |
| Boris Vaňka | 23 | 1 | <NA> | 0 | <NA> |
| Igor Buinevici | 23 | 1 | <NA> | 0 | <NA> |
| Kateřina Žáková | 23 | 1 | <NA> | 0 | <NA> |
| Kateřina „Katka“ Lusková | 23 | 1 | <NA> | 0 | <NA> |
| Klára Zbončáková | 23 | 1 | <NA> | 0 | <NA> |
| Petr Mareš | 23 | 1 | <NA> | 0 | <NA> |
| Helena Hájková | 29 | 1 | <NA> | 0 | <NA> |
| Lukáš Vyskočil (Laskavý chachar) | 29 | 1 | <NA> | 0 | <NA> |
| Pavel Cahlík | 29 | 1 | <NA> | 0 | <NA> |
| Petr Schwank | 29 | 1 | <NA> | 0 | <NA> |
| Michelle Losekoot | 33 | 1 | 4 | 6 | -29 |
| José Kadlec | 33 | 1 | <NA> | 0 | <NA> |
| Martina Špatná | 33 | 1 | <NA> | 0 | <NA> |
| Michal Plichta | 33 | 1 | <NA> | 0 | <NA> |
| Pavla Petrů | 33 | 1 | <NA> | 0 | <NA> |
| Tomáš Černovský | 33 | 1 | <NA> | 0 | <NA> |
| Josef Kadlec | <NA> | 0 | 3 | 7 | <NA> |
| Lenka Rožnovská | <NA> | 0 | 6 | 5 | <NA> |
| Petra Drahoňovská | <NA> | 0 | 10 | 2 | <NA> |
| Kristýna Szkanderová | <NA> | 0 | 11 | 2 | <NA> |
| Jan Bartoš | <NA> | 0 | 12 | 1 | <NA> |
| Lucie Václavková | <NA> | 0 | 13 | 1 | <NA> |
| Pavlína Louženská | <NA> | 0 | 13 | 1 | <NA> |
| Petr Hovorka | <NA> | 0 | 13 | 1 | <NA> |
| Jiří Jambor (a agentura Future Sales) | <NA> | 0 | 16 | 1 | <NA> |
| Ondřej Merta | <NA> | 0 | 16 | 1 | <NA> |
| Polina Ševčíková | <NA> | 0 | 16 | 1 | <NA> |
| Tomáš Vachuda | <NA> | 0 | 16 | 1 | <NA> |
| Vojtěch Strnad | <NA> | 0 | 16 | 1 | <NA> |
Matched product set: claude-sonnet, gemini, openai. Perplexity is search-on only, so the raw arms would conflate enabling search with adding a whole product — the all-products enabled list is its own section below. Places, not ranks: no arm-specific Borda tally exists, so each arm is ordered by ballots then best position, and names with identical evidence share a place so a shift cannot be a tie breaking differently. Positive shift means the name sits higher with search enabled; a blank means it appears on one arm's list only, which is itself the finding._ _Pooled from linkedin-expert rung 1 (54 ballots). One election is one geography, which is how borda.csv keys its own tally — but a rung-2 ask is narrower than a rung-1 one and two intents ask two questions, so more than one family here means this list answers more than one question. Scope the run, or read it as pooled.
| ballots | products | best_rank | mean_rank | |
|---|---|---|---|---|
| name | ||||
| Jana Bártíková | 14 | 3 | 1.0 | 1.83 |
| Míša Chramostová | 10 | 4 | 1.0 | 2.78 |
| Radka Legerská | 9 | 3 | 1.0 | 3.11 |
| Martin Hošek | 7 | 4 | 1.0 | 1.83 |
| Katka Lusková | 6 | 3 | 1.0 | 3.00 |
| Sergej Pavljuk | 5 | 3 | 2.0 | 2.00 |
| Lenka Kazdová | 4 | 1 | 1.0 | 3.00 |
| Jiří Jambor | 4 | 3 | 2.0 | 2.67 |
| Barbora Křížová | 4 | 2 | 2.0 | 4.25 |
| Klára Zbončáková | 4 | 2 | 3.0 | 3.75 |
| Miloš Křepelka | 3 | 2 | 1.0 | 1.00 |
| Martin Schwarz | 3 | 2 | 2.0 | 3.67 |
| Jan Kyselý | 3 | 2 | 3.0 | 3.00 |
| Hana Kavanová | 3 | 1 | 3.0 | 3.50 |
| Adam Zbiejczuk | 3 | 2 | 3.0 | 4.33 |
| Vojtěch Forejtek | 2 | 1 | 1.0 | 1.50 |
| Tereza Kubíčková | 2 | 2 | 3.0 | 3.00 |
| David Kolář | 2 | 1 | 3.0 | 3.50 |
| Eliška Vyhnánková | 2 | 2 | 4.0 | 4.00 |
| Ilona Smelík | 2 | 1 | 4.0 | 4.00 |
| Jarmila Vindušková | 2 | 1 | 4.0 | 4.00 |
| Alžběta Fridrichová | 2 | 2 | 5.0 | 5.00 |
| Lucie Zitterbartová | 1 | 1 | 1.0 | 1.00 |
| Marie Zámečníková | 1 | 1 | 1.0 | 1.00 |
| Michaela (Míša) Chramostová | 1 | 1 | 1.0 | 1.00 |
| Jan Tegze | 1 | 1 | 2.0 | 2.00 |
| Radek Váňa | 1 | 1 | 2.0 | 2.00 |
| Barbora Křivová | 1 | 1 | 3.0 | 3.00 |
| Míša (Michaela) Chramostová | 1 | 1 | 3.0 | 3.00 |
| Tereza (LinkedIn konzultace a personal branding) | 1 | 1 | 3.0 | 3.00 |
| Boris Vaňka | 1 | 1 | 4.0 | 4.00 |
| Igor Buinevici | 1 | 1 | 4.0 | 4.00 |
| Kateřina Žáková | 1 | 1 | 4.0 | 4.00 |
| Kateřina „Katka“ Lusková | 1 | 1 | 4.0 | 4.00 |
| Petr Mareš | 1 | 1 | 4.0 | 4.00 |
| Helena Hájková | 1 | 1 | 5.0 | 5.00 |
| Lenka Kosmatová | 1 | 1 | 5.0 | 5.00 |
| Lukáš Vyskočil (Laskavý chachar) | 1 | 1 | 5.0 | 5.00 |
| Marie Krátká | 1 | 1 | 5.0 | 5.00 |
| Martin Kadlec | 1 | 1 | 5.0 | 5.00 |
| Pavel Cahlík | 1 | 1 | 5.0 | 5.00 |
| Petr Schwank | 1 | 1 | 5.0 | 5.00 |
| Šárka Risch | 1 | 1 | 5.0 | 5.00 |
| José Kadlec | 1 | 1 | ||
| Martina Špatná | 1 | 1 | ||
| Michal Plichta | 1 | 1 | ||
| Michelle Losekoot | 1 | 1 | ||
| Ondřej Souček | 1 | 1 | ||
| Pavla Petrů | 1 | 1 | ||
| Tomáš Černovský | 1 | 1 |
Every product that answered with search enabled, Perplexity included — roadmap §7 asks for this beside the matched contrast, because excluding a search-only product can change the leader. Mentions, not a ranking: borda.csv carries no search column, so no arm-specific tally exists.
| enabled place | enabled ballots | disabled place | disabled ballots | place shift | |
|---|---|---|---|---|---|
| name | |||||
| Ivana Brutenič | 1 | 15 | 5 | 4 | 4 |
| Marta Pálešová | 2 | 14 | <NA> | 0 | <NA> |
| Kristína Cichý Kováčiková | 3 | 12 | <NA> | 0 | <NA> |
| Miki Plichta | 4 | 11 | 1 | 6 | -3 |
| Lukáš Bednár | 5 | 7 | <NA> | 0 | <NA> |
| Daniel Výcha | 6 | 5 | <NA> | 0 | <NA> |
| Lucia Pekaríková | 7 | 5 | <NA> | 0 | <NA> |
| Oliver Jakubík | 8 | 4 | 9 | 3 | 1 |
| Sergej Pavljuk | 9 | 3 | <NA> | 0 | <NA> |
| Tomáš Kromka | 10 | 3 | <NA> | 0 | <NA> |
| Ján Stareček | 11 | 2 | <NA> | 0 | <NA> |
| Martin Volek | 12 | 2 | 13 | 1 | 1 |
| Katarína Košarišťanová | 12 | 2 | <NA> | 0 | <NA> |
| Kristína Kubíková | 14 | 1 | <NA> | 0 | <NA> |
| Marta Palešová | 15 | 1 | <NA> | 0 | <NA> |
| Michal Gula | 16 | 1 | <NA> | 0 | <NA> |
| Anna Heveryová | 17 | 1 | <NA> | 0 | <NA> |
| Hana Jech | 17 | 1 | <NA> | 0 | <NA> |
| Juraj Zaťko | 17 | 1 | <NA> | 0 | <NA> |
| Karin Zvalová | 17 | 1 | <NA> | 0 | <NA> |
| Lucia Pekáriková | 17 | 1 | <NA> | 0 | <NA> |
| Michal Tramita | 17 | 1 | <NA> | 0 | <NA> |
| Simona Mištíková | 17 | 1 | <NA> | 0 | <NA> |
| Veronika Hesková | 17 | 1 | <NA> | 0 | <NA> |
| Michal Ukropec | 25 | 1 | <NA> | 0 | <NA> |
| Pavol Škorpík | 25 | 1 | <NA> | 0 | <NA> |
| Ľubomíra Mardiaková | 25 | 1 | <NA> | 0 | <NA> |
| Peter Šebo | <NA> | 0 | 2 | 5 | <NA> |
| Filip Kuna | <NA> | 0 | 3 | 4 | <NA> |
| Peter Chodelka | <NA> | 0 | 3 | 4 | <NA> |
| Kristína Cichý Kováčová | <NA> | 0 | 6 | 4 | <NA> |
| Tony Dúbravec | <NA> | 0 | 6 | 4 | <NA> |
| Michal Pastier | <NA> | 0 | 8 | 4 | <NA> |
| Juraj Sasko | <NA> | 0 | 10 | 3 | <NA> |
| Andrea Vadkerti | <NA> | 0 | 11 | 2 | <NA> |
| Marek Šulík | <NA> | 0 | 12 | 2 | <NA> |
| Michal Kopecký | <NA> | 0 | 13 | 1 | <NA> |
| Peter Strcula | <NA> | 0 | 13 | 1 | <NA> |
| Ivana Brutenič a Oliver Brutenič | <NA> | 0 | 16 | 1 | <NA> |
| Jaroslav Dodok | <NA> | 0 | 16 | 1 | <NA> |
| Peter Murár | <NA> | 0 | 16 | 1 | <NA> |
| Tomáš Török | <NA> | 0 | 16 | 1 | <NA> |
| Gabriel Tóth | <NA> | 0 | 20 | 1 | <NA> |
| Kika Nevedelová | <NA> | 0 | 20 | 1 | <NA> |
| Nina Krovinová | <NA> | 0 | 22 | 1 | <NA> |
| Roland Vojkovský | <NA> | 0 | 22 | 1 | <NA> |
| Peter Sámel | <NA> | 0 | 24 | 1 | <NA> |
| Sergej Danilov | <NA> | 0 | 24 | 1 | <NA> |
| Kristína Cichý Brutenič | <NA> | 0 | 26 | 1 | <NA> |
| Martin Hoško | <NA> | 0 | 26 | 1 | <NA> |
| Slavomír Bača | <NA> | 0 | 26 | 1 | <NA> |
Matched product set: claude-sonnet, gemini, openai. Perplexity is search-on only, so the raw arms would conflate enabling search with adding a whole product — the all-products enabled list is its own section below. Places, not ranks: no arm-specific Borda tally exists, so each arm is ordered by ballots then best position, and names with identical evidence share a place so a shift cannot be a tie breaking differently. Positive shift means the name sits higher with search enabled; a blank means it appears on one arm's list only, which is itself the finding._ _Pooled from linkedin-expert rung 1 (54 ballots). One election is one geography, which is how borda.csv keys its own tally — but a rung-2 ask is narrower than a rung-1 one and two intents ask two questions, so more than one family here means this list answers more than one question. Scope the run, or read it as pooled.
| ballots | products | best_rank | mean_rank | |
|---|---|---|---|---|
| name | ||||
| Marta Pálešová | 20 | 4 | 1.0 | 1.79 |
| Ivana Brutenič | 15 | 3 | 1.0 | 1.62 |
| Kristína Cichý Kováčiková | 13 | 4 | 1.0 | 2.10 |
| Miki Plichta | 11 | 3 | 1.0 | 1.78 |
| Lukáš Bednár | 9 | 4 | 2.0 | 3.88 |
| Oliver Jakubík | 5 | 3 | 2.0 | 3.20 |
| Michal Tramita | 5 | 2 | 2.0 | 3.40 |
| Daniel Výcha | 5 | 1 | 2.0 | 3.60 |
| Lucia Pekaríková | 5 | 1 | 3.0 | 3.60 |
| Sergej Pavljuk | 3 | 1 | 3.0 | 3.50 |
| Dávid Macej | 3 | 1 | 3.0 | 4.00 |
| Martin Volek | 3 | 3 | 4.0 | 4.33 |
| Tomáš Kromka | 3 | 1 | 5.0 | 5.00 |
| Zuzana Ďuricová | 2 | 1 | 2.0 | 3.50 |
| Ján Stareček | 2 | 1 | 3.0 | 3.00 |
| Pavol Škorpík | 2 | 2 | 4.0 | 4.00 |
| Adam Stečák | 2 | 1 | 4.0 | 4.50 |
| Katarína Košarišťanová | 2 | 1 | 4.0 | 4.50 |
| Monika Filipová | 2 | 1 | 5.0 | 5.00 |
| Kristína Kubíková | 1 | 1 | 1.0 | 1.00 |
| Noema Pšenicová | 1 | 1 | 2.0 | 2.00 |
| Marta Palesova | 1 | 1 | 3.0 | 3.00 |
| Marta Palešová | 1 | 1 | 3.0 | 3.00 |
| Miroslav Reiter | 1 | 1 | 3.0 | 3.00 |
| Lucia Pekariková | 1 | 1 | 4.0 | 4.00 |
| Michal Gula | 1 | 1 | 4.0 | 4.00 |
| Taťána Mráziková | 1 | 1 | 4.0 | 4.00 |
| Anna Heveryová | 1 | 1 | 5.0 | 5.00 |
| Hana Jech | 1 | 1 | 5.0 | 5.00 |
| Juraj Zaťko | 1 | 1 | 5.0 | 5.00 |
| Karin Zvalová | 1 | 1 | 5.0 | 5.00 |
| Lucia Pekáriková | 1 | 1 | 5.0 | 5.00 |
| Simona Mištíková | 1 | 1 | 5.0 | 5.00 |
| Veronika Hesková | 1 | 1 | 5.0 | 5.00 |
| Michal Ukropec | 1 | 1 | ||
| Ľubomíra Mardiaková | 1 | 1 |
Every product that answered with search enabled, Perplexity included — roadmap §7 asks for this beside the matched contrast, because excluding a search-only product can change the leader. Mentions, not a ranking: borda.csv carries no search column, so no arm-specific tally exists.
| enabled place | enabled ballots | disabled place | disabled ballots | place shift | |
|---|---|---|---|---|---|
| name | |||||
| Richard van der Blom | 1 | 9 | 2 | 7 | 1 |
| Justin Welsh | 2 | 7 | 1 | 8 | -1 |
| Jasmin Alić | 3 | 5 | 4 | 3 | 1 |
| Lara Acosta | 4 | 4 | 3 | 4 | -1 |
| Amelia Sordell | 5 | 3 | 13 | 1 | 8 |
| Dorie Clark | 5 | 3 | <NA> | 0 | <NA> |
| Joshua B. Lee | 7 | 3 | <NA> | 0 | <NA> |
| Andre Santos | 8 | 2 | <NA> | 0 | <NA> |
| William Arruda | 9 | 2 | <NA> | 0 | <NA> |
| Melonie Dodaro | 10 | 2 | <NA> | 0 | <NA> |
| Daniel Disney | 11 | 2 | 6 | 3 | -5 |
| Jasmin Alić (Jay) | 12 | 1 | <NA> | 0 | <NA> |
| André Santos | 13 | 1 | <NA> | 0 | <NA> |
| Fredrik Fornes | 13 | 1 | <NA> | 0 | <NA> |
| Sam G. Winsbury | 15 | 1 | <NA> | 0 | <NA> |
| Michaela Alexis | 16 | 1 | 8 | 1 | -8 |
| Blossom Affia | 16 | 1 | <NA> | 0 | <NA> |
| Matt Barker | 18 | 1 | 11 | 1 | -7 |
| Judi Fox | 18 | 1 | <NA> | 0 | <NA> |
| Louise Brogan | 18 | 1 | <NA> | 0 | <NA> |
| Goldie Chan | 21 | 1 | 7 | 2 | -14 |
| Jasmin Alic | <NA> | 0 | 4 | 3 | <NA> |
| Lea Turner | <NA> | 0 | 9 | 1 | <NA> |
| Michelle J Raymond | <NA> | 0 | 9 | 1 | <NA> |
| Salina Yeung | <NA> | 0 | 11 | 1 | <NA> |
| AJ Wilcox | <NA> | 0 | 13 | 1 | <NA> |
| Amelia Sordell (Klowt) | <NA> | 0 | 13 | 1 | <NA> |
| Brynne Tillman | <NA> | 0 | 13 | 1 | <NA> |
| Josh Fechter | <NA> | 0 | 13 | 1 | <NA> |
| Richard van der Blom (Just Connecting) | <NA> | 0 | 13 | 1 | <NA> |
| Sam Browne | <NA> | 0 | 13 | 1 | <NA> |
| Sam Browne (SocialSam) | <NA> | 0 | 13 | 1 | <NA> |
Matched product set: claude-sonnet, gemini, openai. Perplexity is search-on only, so the raw arms would conflate enabling search with adding a whole product — the all-products enabled list is its own section below. Places, not ranks: no arm-specific Borda tally exists, so each arm is ordered by ballots then best position, and names with identical evidence share a place so a shift cannot be a tie breaking differently. Positive shift means the name sits higher with search enabled; a blank means it appears on one arm's list only, which is itself the finding._ _Pooled from linkedin-expert rung 1 (27 ballots). One election is one geography, which is how borda.csv keys its own tally — but a rung-2 ask is narrower than a rung-1 one and two intents ask two questions, so more than one family here means this list answers more than one question. Scope the run, or read it as pooled.
| ballots | products | best_rank | mean_rank | |
|---|---|---|---|---|
| name | ||||
| Richard van der Blom | 10 | 4 | 1.0 | 1.67 |
| Justin Welsh | 10 | 4 | 1.0 | 1.88 |
| Jasmin Alić | 8 | 3 | 2.0 | 2.88 |
| Lara Acosta | 7 | 3 | 2.0 | 3.17 |
| Dorie Clark | 4 | 2 | 2.0 | 3.50 |
| Amelia Sordell | 3 | 3 | 2.0 | 3.67 |
| Goldie Chan | 3 | 2 | 4.0 | 4.00 |
| Joshua B. Lee | 3 | 2 | 4.0 | 4.50 |
| Andre Santos | 2 | 2 | 1.0 | 2.00 |
| William Arruda | 2 | 1 | 3.0 | 4.00 |
| Melonie Dodaro | 2 | 1 | 4.0 | 4.00 |
| Bhavik Sarkhedi | 2 | 1 | 5.0 | 5.00 |
| Daniel Disney | 2 | 1 | 5.0 | 5.00 |
| Jasmin Alić (Jay) | 1 | 1 | 1.0 | 1.00 |
| André Santos | 1 | 1 | 2.0 | 2.00 |
| Fredrik Fornes | 1 | 1 | 2.0 | 2.00 |
| Sam G. Winsbury | 1 | 1 | 3.0 | 3.00 |
| Blossom Affia | 1 | 1 | 4.0 | 4.00 |
| Michaela Alexis | 1 | 1 | 4.0 | 4.00 |
| Judi Fox | 1 | 1 | 5.0 | 5.00 |
| Louise Brogan | 1 | 1 | 5.0 | 5.00 |
| Matt Barker | 1 | 1 | 5.0 | 5.00 |
| Arman Harutyunyan | 1 | 1 | ||
| Ashraf Malik | 1 | 1 | ||
| Carol Kaemmerer | 1 | 1 | ||
| Hanna Larsson | 1 | 1 | ||
| Molly Stovold | 1 | 1 |
Every product that answered with search enabled, Perplexity included — roadmap §7 asks for this beside the matched contrast, because excluding a search-only product can change the leader. Mentions, not a ranking: borda.csv carries no search column, so no arm-specific tally exists.
| ballots | share | |
|---|---|---|
| names in answer | ||
| 0 | 21 | 15.6% |
| 2 | 2 | 1.5% |
| 4 | 1 | 0.7% |
| 5 | 88 | 65.2% |
| 6 | 11 | 8.1% |
| 7 | 5 | 3.7% |
| 8 | 3 | 2.2% |
| 9 | 1 | 0.7% |
| 10 | 3 | 2.2% |
A model asked for five returns three, or seven. This distribution is what the truncated-ballot decision has to be made against; the over-length count is the honourable-mention tail.
| candidacies | share | |
|---|---|---|
| placement | ||
| enumerated | 487 | 79.6% |
| narrative | 91 | 14.9% |
| honourable_mention | 34 | 5.6% |
enumerated is a numbered slot and carries a real position. narrative and honourable_mention do not, and scoring them as if they did is the quiet error a Borda tally is most sensitive to.
| response_mode | named_people | generic_advice | refusal | clarifying_question | n | |
|---|---|---|---|---|---|---|
| subject | search_mode | |||||
| claude-sonnet | off | 13.3% | 86.7% | 0.0% | 0.0% | 15 |
| on | 100.0% | 0.0% | 0.0% | 0.0% | 15 | |
| deepseek | off | 100.0% | 0.0% | 0.0% | 0.0% | 15 |
| gemini | off | 100.0% | 0.0% | 0.0% | 0.0% | 15 |
| on | 100.0% | 0.0% | 0.0% | 0.0% | 15 | |
| kimi | off | 46.7% | 46.7% | 6.7% | 0.0% | 15 |
| openai | off | 100.0% | 0.0% | 0.0% | 0.0% | 15 |
| on | 100.0% | 0.0% | 0.0% | 0.0% | 15 | |
| perplexity | on | 100.0% | 0.0% | 0.0% | 0.0% | 15 |
Over the seated ballots only; the excluded families are not in this denominator.
every completed run in this directory is a ballot
| search arms | ballots | names | slots | unique names | unique share | consensus capture | ||
|---|---|---|---|---|---|---|---|---|
| election | product | |||||||
| czechia | kimi | off only | 4 | 10 | 13 | 7 | 70.0% | 46.2% |
| deepseek | off only | 6 | 19 | 30 | 13 | 68.4% | 40.0% | |
| perplexity | on only | 6 | 17 | 33 | 11 | 64.7% | 33.3% | |
| gemini | on+off | 12 | 30 | 61 | 15 | 50.0% | 59.0% | |
| claude-sonnet | on only | 6 | 19 | 31 | 7 | 36.8% | 61.3% | |
| openai | on+off | 12 | 21 | 60 | 5 | 23.8% | 76.7% | |
| slovakia | deepseek | off only | 6 | 24 | 31 | 21 | 87.5% | 25.8% |
| gemini | on+off | 12 | 24 | 66 | 16 | 66.7% | 59.1% | |
| openai | on+off | 12 | 25 | 61 | 13 | 52.0% | 41.0% | |
| perplexity | on only | 6 | 15 | 29 | 7 | 46.7% | 41.4% | |
| claude-sonnet | on only | 6 | 14 | 33 | 6 | 42.9% | 66.7% | |
| world | perplexity | on only | 3 | 12 | 20 | 6 | 50.0% | 65.0% |
| claude-sonnet | on+off | 5 | 12 | 25 | 5 | 41.7% | 56.0% | |
| deepseek | off only | 3 | 15 | 29 | 5 | 33.3% | 58.6% | |
| gemini | on+off | 6 | 12 | 31 | 4 | 33.3% | 83.9% | |
| openai | on+off | 6 | 16 | 39 | 5 | 31.2% | 64.1% | |
| kimi | off only | 3 | 9 | 19 | 0 | 0.0% | 94.7% |
Distinctive is not a verdict. A name only one product offers may be a name only one product knows, and a product that captures the consensus may be naming the same five famous people as everyone else. unique share is over this product's resolved names; consensus capture is over its ballot slots, against the quorum floor of a tally computed without it — scored against the full tally, a product that repeats one name would push that name over the floor itself and be paid for its own consensus. Read both against ballots: a share over four ballots is not the same fact as one over twenty-one. Search arms pool per product here, and the arms each product ran are in the first column.
| rbo | jaccard | shared names | |||
|---|---|---|---|---|---|
| election | product | and | |||
| czechia | gemini | openai | 0.453 | 0.244 | 10 |
| claude-sonnet | gemini | 0.414 | 0.256 | 10 | |
| openai | 0.226 | 0.250 | 8 | ||
| deepseek | openai | 0.221 | 0.111 | 4 | |
| openai | perplexity | 0.125 | 0.152 | 5 | |
| gemini | perplexity | 0.119 | 0.068 | 3 | |
| claude-sonnet | perplexity | 0.108 | 0.091 | 3 | |
| deepseek | kimi | 0.045 | 0.115 | 3 | |
| claude-sonnet | deepseek | 0.037 | 0.027 | 1 | |
| deepseek | gemini | 0.029 | 0.065 | 3 | |
| claude-sonnet | kimi | 0.000 | 0.036 | 1 | |
| deepseek | perplexity | 0.000 | 0.000 | 0 | |
| gemini | kimi | 0.000 | 0.053 | 2 | |
| kimi | openai | 0.000 | 0.033 | 1 | |
| perplexity | 0.000 | 0.000 | 0 | ||
| slovakia | claude-sonnet | gemini | 0.473 | 0.188 | 6 |
| gemini | openai | 0.381 | 0.167 | 7 | |
| claude-sonnet | openai | 0.352 | 0.219 | 7 | |
| perplexity | 0.342 | 0.261 | 6 | ||
| openai | perplexity | 0.332 | 0.176 | 6 | |
| gemini | perplexity | 0.210 | 0.147 | 5 | |
| deepseek | openai | 0.082 | 0.065 | 3 | |
| claude-sonnet | deepseek | 0.000 | 0.000 | 0 | |
| deepseek | gemini | 0.000 | 0.000 | 0 | |
| perplexity | 0.000 | 0.000 | 0 | ||
| world | kimi | perplexity | 0.811 | 0.312 | 5 |
| gemini | kimi | 0.645 | 0.312 | 5 | |
| openai | 0.643 | 0.400 | 8 | ||
| kimi | openai | 0.637 | 0.389 | 7 | |
| deepseek | kimi | 0.616 | 0.412 | 7 | |
| openai | 0.546 | 0.240 | 6 | ||
| claude-sonnet | deepseek | 0.532 | 0.174 | 4 | |
| openai | 0.509 | 0.167 | 4 | ||
| gemini | perplexity | 0.504 | 0.200 | 4 | |
| openai | perplexity | 0.496 | 0.217 | 5 | |
| deepseek | perplexity | 0.435 | 0.174 | 4 | |
| gemini | 0.389 | 0.125 | 3 | ||
| claude-sonnet | gemini | 0.377 | 0.200 | 4 | |
| perplexity | 0.364 | 0.200 | 4 | ||
| kimi | 0.359 | 0.235 | 4 |
Two measures because they disagree in the informative case. jaccard is over resolved name sets — did they name the same people at all. rbo is rank-biased overlap (p=0.9) over each product's own Borda ranking — did they put them in the same order, weighted towards the head. Kendall's τ is not here on purpose: it is defined over the items two rankings share, so on top-heavy lists with disjoint tails it would score the intersection and discard the disagreement it was called to measure.
| top-5 shared | rbo | jaccard | shared names | only in the first ask's table | only in the second ask's table | ||
|---|---|---|---|---|---|---|---|
| election | asks | ||||||
| czechia | en vs cs | 3 | 0.593 | 0.325 | 26 | Miloš Křepelka; Jiří Jambor; Jan Tegze; Klára Zbončáková; Barbora Křížová; Petra Drahoňovská; Jan Kyselý; Ilona Smelík; Jarmila Vindušková; Tereza Kubíčková | Katka Lusková; Vojtěch Forejtek; Hana Kavanová; Daniel Gamrot; Martin Schwarz |
| slovakia | en vs sk | 4 | 0.673 | 0.286 | 22 | Filip Kuna; Dávid Macej; Adam Stečák; Pavol Škorpík; Katarína Košarišťanová; Sergej Pavljuk | Michal Tramita; Peter Chodelka; Andrea Vadkerti; Ján Stareček; Lucia Mokráňová; Martin Volek; Marek Šulík; Monika Filipová; Juraj Sasko; Tomáš Kromka |
The language of the ask is the treatment and the products pool. Neither ask is the right one: a name in one ask's table and not the other's is a name that ask surfaced, and it may sit in the other's long tail rather than be absent from it — the lists are drawn at the quorum floor. Each arm holds about twenty ballots, so a change at the head is one or two answers moving.
| name | near-duplicate of | evidence | mentions | mentions of the other | |
|---|---|---|---|---|---|
| 0 | Ivana Brutenič | Ivana Brutenič a Oliver Brutenič | shares a surname token | 19 | 1 |
| 1 | Ivana Brutenič | Kristína Cichý Brutenič | shares a surname token | 19 | 1 |
| 2 | Miki Plichta | Michal Plichta | shares a surname token | 17 | 1 |
| 3 | Josef Kadlec | Martin Kadlec | shares a surname token | 8 | 1 |
| 4 | Ivana Brutenič a Oliver Brutenič | Kristína Cichý Brutenič | shares a surname token | 1 | 1 |
Nothing here is merged: a false merge inflates one candidate and erases another, which in a ranking is worse than a false split. Two rows on one /in/ profile are a certain split; a shared surname token is a candidate one for a human to decide.
| election | borda | n_ballots | n_products | best_rank | |
|---|---|---|---|---|---|
| person | |||||
| Jiří Jambor | czechia | 0.236 | 6 | 3 | 2 |
| Josef Kadlec | czechia | 0.208 | 8 | 2 | 2 |
| Martin Kadlec | czechia | 0.028 | 1 | 1 | 5 |
| Jiří Jambor (a agentura Future Sales) | czechia | 0.000 | 1 | 1 | <NA> |
| Michal Plichta | czechia | 0.000 | 1 | 1 | <NA> |
| Ivana Brutenič | slovakia | 1.483 | 19 | 3 | 1 |
| Miki Plichta | slovakia | 1.017 | 17 | 3 | 1 |
| Ivana Brutenič a Oliver Brutenič | slovakia | 0.067 | 1 | 1 | 2 |
| Kristína Cichý Brutenič | slovakia | 0.000 | 1 | 1 | <NA> |
| Amelia Sordell | world | 0.200 | 5 | 4 | 2 |
| Amelia Sordell (Klowt) | world | 0.000 | 1 | 1 | <NA> |
| Sam Browne | world | 0.000 | 1 | 1 | <NA> |
| Sam Browne (SocialSam) | world | 0.000 | 1 | 1 | <NA> |
| n_mentions | best_position | summary | links | subjects | rungs | |
|---|---|---|---|---|---|---|
| name | ||||||
| Jana Bártíková | 25 | 1 | Founder of Linked Akademie; teaches LinkedIn strategy for B2B business, personal branding, and customer acquisition in Czech Republic. | cz.linkedin.com · linkedakademie.cz · linkedacademy.eu | claude-sonnet×3, gemini×10, openai×12 | r1×25 |
| Richard van der Blom | 24 | 1 | LinkedIn strategist with nearly two decades of experience, known for rigorous algorithm research and advising 1,000+ B2B teams including Salesforce, PwC and Capgemini. | linkedin.com · linkedin.com · richardvanderblom.com | claude-sonnet×5, deepseek×3, gemini×6, kimi×3, openai×6, perplexity×1 | r1×24 |
| Justin Welsh | 23 | 1 | Solo creator teaching LinkedIn audience growth and personal branding through The LinkedIn Operating System course. | learn.justinwelsh.me · learn.justinwelsh.me | claude-sonnet×5, deepseek×2, gemini×5, kimi×3, openai×5, perplexity×3 | r1×23 |
| Marta Pálešová | 22 | 1 | LinkedIn consultant and trainer specializing in personal branding, profile optimization, and lead generation in Slovakia. | sk.linkedin.com · sk.linkedin.com · martapalesova.com | claude-sonnet×4, gemini×5, openai×6, perplexity×7 | r1×22 |
| Ivana Brutenič | 19 | 1 | Co-founder of SUNDAYFLIES, LinkedIn consultancy specializing in employee advocacy and B2B strategy in Slovakia. | sk.linkedin.com · sundayflies.com · sundayflies.com | claude-sonnet×5, gemini×10, openai×4 | r1×19 |
| Jasmin Alić | 18 | 1 | Fortune 500 copywriter, ranked #1 LinkedIn expert globally by Favikon, coaches founders and executives, runs 'Link Up' community, known for commenting strategy and text-only copy. | ba.linkedin.com | claude-sonnet×1, gemini×6, kimi×3, openai×5, perplexity×3 | r1×18 |
| Miki Plichta | 17 | 1 | LinkedIn Storyteller specializing in personal branding and content creation for professionals and corporate clients in Slovakia. | sk.linkedin.com · miki.sk · miki.sk | claude-sonnet×1, gemini×12, openai×4 | r1×17 |
| Lara Acosta | 16 | 2 | LinkedIn personal branding expert running LA Digital agency; teaches framework for rapid founder audience growth. | uk.linkedin.com | deepseek×2, gemini×5, kimi×3, openai×3, perplexity×3 | r1×16 |
| Sergej Pavljuk | 13 | 2 | Founder/co-founder of StoryMatters, Europe's largest LinkedIn communications agency; B2B marketing and social selling strategist. | cz.linkedin.com | claude-sonnet×3, gemini×9, openai×1 | r1×13 |
| Eliška Vyhnánková | 13 | 2 | Czech social media strategist and trainer; co-author of Jak na sítě; LinkedIn and personal branding specialist. | cz.linkedin.com | deepseek×2, gemini×3, kimi×1, openai×7 | r1×13 |
| Kristína Cichý Kováčiková | 13 | 1 | Co-founder of SUNDAYFLIES, LinkedIn specialist in employee advocacy and employer branding, co-author of The Magic of Employee Influence. | sk.linkedin.com | claude-sonnet×4, gemini×6, openai×2, perplexity×1 | r1×13 |
| Michaela Chramostová | 12 | 1 | LinkedIn Personal Brand Strategist helping companies, entrepreneurs, and experts build their personal brand, relationships, and business on LinkedIn; 16 years B2B experience, 1000+ professionals trained. | cz.linkedin.com | claude-sonnet×3, gemini×1, openai×2, perplexity×6 | r1×12 |
| Michelle Losekoot | 12 | 1 | Content strategist and personal branding specialist helping leaders build authentic LinkedIn presence; based in Prague, Czechia. | deepseek×5, gemini×1, openai×6 | r1×12 | |
| Goldie Chan | 10 | 4 | LinkedIn video pioneer and personal branding strategist for executives, especially introverts. | deepseek×3, kimi×2, openai×3, perplexity×2 | r1×10 | |
| Michal Pastier | 10 | 1 | Respected Slovak brand strategist, author and speaker specializing in personal branding and thought leadership. | deepseek×6, openai×4 | r1×10 | |
| Lukáš Bednár | 9 | 2 | Digital marketing strategist and LinkedIn trainer at DigiSkilly.sk; Google Certified Trainer focusing on B2B strategy and personal branding. | sk.linkedin.com · digiskilly.sk | claude-sonnet×3, gemini×2, openai×2, perplexity×2 | r1×9 |
| Martin Hošek | 9 | 1 | Czech LinkedIn consultant and personal branding strategist; founder of LiSummit conference and MHOSEK.CZ. | mhosek.cz | claude-sonnet×2, gemini×4, openai×2, perplexity×1 | r1×9 |
| Radka Legerská | 9 | 1 | B2B marketing strategist specializing in LinkedIn strategy, employer branding, executive ghostwriting, and corporate training. | cz.linkedin.com · kreativnicesko.cz · 2024.marketingfestival.cz | claude-sonnet×2, gemini×1, openai×6 | r1×9 |
| Oliver Jakubík | 8 | 2 | HR marketing and employer branding consultant in Slovakia; trains personal and corporate branding since 2008. | oliverjakubik.com | claude-sonnet×3, gemini×3, openai×1, perplexity×1 | r1×8 |
| Adam Zbiejczuk | 8 | 2 | Czech social media strategist and lecturer specializing in LinkedIn, co-owner of Influencer.cz. | claude-sonnet×2, deepseek×2, gemini×1, kimi×3 | r1×8 | |
| Josef Kadlec | 8 | 2 | Co-founder of GoodCall and Recruitment Academy, LinkedIn expert focusing on recruitment, social selling, and personal branding in Czechia. | gemini×4, openai×4 | r1×8 | |
| Lucia Pekaríková | 7 | 3 | CEO of KOMUNIKA PR, personal branding strategist with 18+ years in PR across Central Europe. | sk.linkedin.com | openai×6, perplexity×1 | r1×7 |
| Katka Lusková | 7 | 1 | Personal brand strategist for CEOs and founders with 17+ years experience; BE THE BRAND founder. | cz.linkedin.com | claude-sonnet×1, openai×3, perplexity×3 | r1×7 |
| Jiří Jambor | 6 | 2 | Co-founder of Future Sales, B2B LinkedIn strategy and social selling agency serving 160+ clients. | cz.linkedin.com | claude-sonnet×1, gemini×3, openai×2 | r1×6 |
| Michaela Alexis | 6 | 2 | LinkedIn trainer and speaker specializing in corporate storytelling, personal branding, and employee advocacy. | ca.linkedin.com | deepseek×3, kimi×1, openai×2 | r1×6 |
| Peter Šebo | 6 | 2 | Founder of PS:Digital, digital strategist specializing in corporate social media, branding, and LinkedIn programs. | linkedin.com · psdigital.sk | deepseek×1, openai×5 | r1×6 |
| Lenka Rožnovská | 5 | 5 | LinkedIn personal branding and copywriting mentor helping professionals build authentic voice and engaging content. | gemini×5 | r1×5 | |
| Viveka von Rosen | 5 | 2 | LinkedIn marketing expert and consultant; author of LinkedIn Marketing: An Hour a Day; co-founder of Vengreso. | deepseek×3, kimi×2 | r1×5 | |
| Michal Tramita | 5 | 2 | LinkedIn strategy consultant at Libellius, specializing in profile optimization and AI integration for business growth. | claude-sonnet×1, perplexity×4 | r1×5 | |
| Daniel Výcha | 5 | 2 | CEO of CoolWriters content agency specializing in copywriting and LinkedIn training for corporations. | claude-sonnet×5 | r1×5 | |
| Daniel Disney | 5 | 4 | LinkedIn social selling trainer and author helping B2B professionals convert visibility into revenue through Sales Navigator expertise. | uk.linkedin.com · linkedin.com | gemini×1, openai×4 | r1×5 |
| Amelia Sordell | 5 | 2 | Founder of Klowt personal branding agency; LinkedIn's Top Voice on Personal Branding UK 2024. | klowt.com · uk.linkedin.com | claude-sonnet×2, gemini×1, kimi×1, openai×1 | r1×5 |
| Kristína Cichý Kováčová | 4 | 3 | B2B marketing and employer branding specialist, runs LinkedIn-focused agency helping companies and HR leaders. | gemini×4 | r1×4 | |
| Tony Dúbravec | 4 | 3 | Digital marketing consultant and podcaster specializing in content strategy and LinkedIn for professionals. | gemini×4 | r1×4 | |
| Peter Chodelka | 4 | 1 | Personal branding and LinkedIn content strategist specializing in audience building for entrepreneurs and B2B professionals. | openai×4 | r1×4 | |
| Martin Volek | 4 | 1 | Zameriava sa na B2B marketing, firemné LinkedIn stratégie, reklamu, predaj a získavanie leadov; školil O2 Business Services, Národnú diaľničnú spoločnosť a ďalších. | linkedin.com | claude-sonnet×1, openai×2, perplexity×1 | r1×4 |
| Lenka Kazdová | 4 | 1 | LinkedIn Strategist specializing in personal branding, visibility, and profile optimization since 2013. | perplexity×4 | r1×4 | |
| Filip Kuna | 4 | 1 | Content marketing practitioner and LinkedIn trainer specializing in employee advocacy and thought leadership in Slovakia. | linkedin.com | gemini×1, openai×3 | r1×4 |
| Klára Zbončáková | 4 | 3 | LinkedIn consultant offering personalized consulting, workshops, mentoring, and account management for personal branding. | gemini×1, perplexity×3 | r1×4 | |
| Barbora Křížová | 4 | 2 | LinkedIn consultant helping entrepreneurs and B2B companies with strategy, content, and lead generation. | cz.linkedin.com | openai×3, perplexity×1 | r1×4 |
| Filip Dřímalka | 4 | 2 | Digital transformation consultant and Digiskills founder focused on digital skills training and LinkedIn strategy. | deepseek×2, kimi×2 | r1×4 | |
| Dorie Clark | 4 | 2 | Author and executive strategist specializing in thought leadership and personal branding; teaches at Duke Fuqua. | claude-sonnet×3, perplexity×1 | r1×4 | |
| Judi Fox | 4 | 5 | LinkedIn coach and speaker specializing in growth strategy, engagement, and content creation. | claude-sonnet×1, deepseek×3 | r1×4 | |
| Joshua B. Lee | 3 | 4 | Best for founder and executive authority; runs StandOut Authority, focuses on positioning and executive content, clients include Gartner and Coursera. | standoutauthority.com | gemini×1, openai×2 | r1×3 |
| Tomáš Kromka | 3 | 5 | Ranked among top 32 LinkedIn creators in Slovakia; LinkedIn trainer, consultant, and management mentor combining coaching background with LinkedIn strategies. | gemini×3 | r1×3 | |
| Petra Drahoňovská | 3 | 3 | Best for personal branding, career positioning, LinkedIn profile and storytelling for professionals; works with individuals rebuilding their professional image. | deepseek×1, openai×2 | r1×3 | |
| Miloš Křepelka | 3 | 1 | LinkedIn Specialist & Sales Strategist who trained over 250 users in a year and created LiDAY, the country's flagship LinkedIn conference. | claude-sonnet×2, gemini×1 | r1×3 | |
| Melonie Dodaro | 3 | 3 | Founder and CEO of Top Dog Social Media; Amazon bestselling author of six books on LinkedIn and social selling, helps convert visibility into leads. | claude-sonnet×2, deepseek×1 | r1×3 | |
| Lucia Mokráňová | 3 | 2 | Communications and personal-branding specialist working with executives and professionals, focused on reputation and LinkedIn storytelling. | deepseek×3 | r1×3 | |
| Juraj Sasko | 3 | 5 | Respected Slovak digital-marketing strategist and co-founder of Visibility; best for companies wanting agency resources and measurable digital distribution. | linkedin.com · visibility.sk | openai×3 | r1×3 |
| Martin Schwarz | 3 | 2 | Nabízí optimalizaci profilů, konzultace, školení, obsahovou tvorbu, ghostwriting a kompletní správu LinkedIn komunikace, specializuje se na IT, technologie a finance. | cz.linkedin.com | gemini×2, openai×1 | r1×3 |
| Jan Tegze | 3 | 2 | Internationally recognised Czech recruiter and author of Full Stack Recruiter, best for recruitment, HR, talent acquisition and thought-leadership positioning. | claude-sonnet×1, openai×2 | r1×3 | |
| Hana Kavanová | 3 | 3 | Specialistka na LinkedIn a osobní branding, nabízí krátké online konzultace, pracuje s jednotlivci i malými podnikateli. | perplexity×3 | r1×3 | |
| Dávid Macej | 3 | 3 | Online Marketing Consultant, LinkedIn Certified Professional, LinkedIn Trainer, and LinkedIn Ads Specialist helping companies with organic and paid LinkedIn presence. | perplexity×3 | r1×3 | |
| Daniel Gamrot | 3 | 2 | Specialista na LinkedIn reklamu a analytiku, zakladatel projektu DigiSemafor, vhodný pro placenou propagaci a výkonnostní kampaně na LinkedInu. | deepseek×3 | r1×3 | |
| Andy Foote | 3 | 3 | Respected LinkedIn ghostwriter who handles writing and positioning for CEOs, VCs, consultants, and senior professionals. | deepseek×3 | r1×3 | |
| Andrea Vadkerti | 3 | 1 | Dlhodobo najvýraznejšie meno v LinkedIn poradenstve na Slovensku, venuje sa osobnej značke, LinkedIn Academy a firemným tréningom. | linkedin.com · andreavadkerti.com | deepseek×1, openai×2 | r1×3 |
| Andre Santos | 3 | 1 | Executive positioning and corporate reputation expert from Brazil; mentors C-level leaders across multiple countries. | favikon.com · br.linkedin.com | gemini×1, openai×2 | r1×3 |
| Jan Kyselý | 3 | 3 | Spoluzakladatel poradenské společnosti FUTURE SALES a autor první české knihy o Social Sellingu, zaměřuje se na B2B marketing a generování leadů. | claude-sonnet×1, gemini×2 | r1×3 | |
| Pavol Škorpík | 2 | 4 | Performance-oriented LinkedIn marketing trainer at wooacademy, ex-marketing director of Pelikán, offers individual LinkedIn training online or in Bratislava. | gemini×1, perplexity×1 | r1×2 | |
| Marek Šulík | 2 | 3 | Digitálny stratég z agentúry Visibility, vhodný pre firmy chcúce zapojiť LinkedIn do širšej marketingovej stratégie. | openai×2 | r1×2 | |
| Matt Barker | 2 | 5 | Premier LinkedIn ghostwriter specializing in B2B lead generation copy; offers digital products, templates, and high-ticket ghostwriting services. | gemini×2 | r1×2 | |
| Monika Filipová | 2 | 5 | Mentorka identity a osobnej značky, využíva LinkedIn ako hlavnú sieť na strategické budovanie osobnej značky cez obsah. | perplexity×2 | r1×2 | |
| Pavlína Louženská | 2 | 5 | Brand strategist and community builder, active in the Czech marketing scene. Helpful for positioning, brand strategy, and content direction. | deepseek×1, openai×1 | r1×2 | |
| Zuzana Ďuricová | 2 | 2 | Focuses on personal branding, PR, social media strategy, and storytelling that builds trust and visibility, with LinkedIn as a key channel. | perplexity×2 | r1×2 | |
| Tereza Kubíčková | 2 | 3 | Works with founders, executives and experts on positioning, messaging, LinkedIn strategy, content and ghostwriting in Czech and English. | cz.linkedin.com | openai×1, perplexity×1 | r1×2 |
| Vojtěch Forejtek | 2 | 1 | Zakladatel StoryMatters.online, B2B LinkedIn agentury, která firmám pomáhá s generováním obchodních schůzek a náborem, buduje reputaci firem a lídrů na LinkedInu, 100+ klientů a 25členný tým. | claude-sonnet×2 | r1×2 | |
| William Arruda | 2 | 3 | Often called the father of personal branding; author of Career Distinction and founder of Reach Personal Branding. | claude-sonnet×2 | r1×2 | |
| Kristýna Szkanderová | 2 | LinkedIn strategist and consultant helping individuals build personal brands and companies establish employee advocacy programs. | gemini×2 | r1×2 | ||
| Marek Prchal | 2 | 4 | Best for founder-led marketing, B2B content strategy, social selling; broader B2B marketing and social media consultant. | deepseek×2 | r1×2 | |
| Katarína Košarišťanová | 2 | 4 | Through Noticing, offers LinkedIn strategy for personal profiles, company pages and ambassador programmes with a B2B sales orientation; achieved fourfold company-page growth and ~€150,000 in sales from an ambassador programme. | gonoticing.com · sk.linkedin.com | openai×2 | r1×2 |
| John Espirian | 2 | 3 | Author of Content DNA; practical and technical expert on profile optimization, company pages and content strategy. | deepseek×1, kimi×1 | r1×2 | |
| Jarmila Vindušková | 2 | 4 | Focuses on corporate and personal 'ambasadoring' on LinkedIn, coordinating multiple employee profiles for unified brand. | perplexity×2 | r1×2 | |
| Jan Řezáč | 2 | 1 | Digital marketing and strategy consultant, founder of House of Řezáč. Strong on credible digital presence and marketing thinking, though not LinkedIn-only. | deepseek×2 | r1×2 | |
| Ilona Smelík | 2 | 4 | LinkedIn Strategy & Content specialist and B2B marketing copywriter at Ili Marketing, focused on content that converts. | perplexity×2 | r1×2 | |
| David Kolář | 2 | 3 | CEO of GetLeads, a Czech agency focused on LinkedIn & B2B marketing generating measurable leads for clients. | claude-sonnet×2 | r1×2 | |
| Bhavik Sarkhedi | 2 | 5 | Content and narrative specialist, ranked #1 in one list, known for crafting compelling LinkedIn profiles and content. | perplexity×2 | r1×2 | |
| Alžběta Fridrichová | 2 | 5 | CEO pražské PR agentury Adison, která nabízí LinkedIn ambasadoring - firemní program pro budování firemní i osobní reputace na LinkedInu. | claude-sonnet×1, gemini×1 | r1×2 | |
| Adam Stečák | 2 | 4 | Specializes in personal branding and marketing for professionals in finance, real estate, and entrepreneurship, teaching people how to become a brand. | perplexity×2 | r1×2 | |
| AJ Wilcox | 2 | Suggested as head of a different shortlist specifically for paid LinkedIn advertising. | deepseek×1, openai×1 | r1×2 | ||
| Ján Stareček | 2 | 3 | Špecialista na dátovú analýzu slovenského LinkedIn z Data on Steroids, tvorca rebríčka TOP 100 tvorcov slovenského LinkedIn, má vlastný podcast. | claude-sonnet×2 | r1×2 | |
| Michal Kopecký | 1 | 1 | najlepšia voľba pre LinkedIn stratégiu, osobnú značku, obsah a získavanie obchodných príležitostí v B2B | openai×1 | r1×1 | |
| Peter Strcula | 1 | 1 | LinkedIn and social-selling specialist with experience training executives, salespeople and B2B teams. | openai×1 | r1×1 | |
| Peter Kolesár | 1 | 1 | LinkedIn stratégia, firemné školenia, osobný branding, predaj a budovanie publika | deepseek×1 | r1×1 | |
| Peter Krajčovič | 1 | 3 | Špecialista na LinkedIn z pohľadu B2B a social sellingu, praktické tréningy pre firmy aj jednotlivcov. | deepseek×1 | r1×1 | |
| Peter Murár | 1 | 2 | Skúsený konzultant a lektor sociálnych sietí vrátane LinkedInu, vhodný na školenie a praktické vedenie. | openai×1 | r1×1 | |
| Peter Mészáros | 1 | 4 | Venuje sa osobnej značke, storytellingu a LinkedIn obsahu, pomáha ľuďom vystupovať pod vlastným menom. | deepseek×1 | r1×1 | |
| Petr Casanova | 1 | 4 | Specialista na osobní branding obecně (nejen LinkedIn), hodí se spíš pro strategickou rovinu značky. | kimi×1 | r1×1 | |
| Peter Sámel | 1 | 5 | Expert na Social Selling, B2B predaj a generovanie leadov cez LinkedIn. | gemini×1 | r1×1 | |
| Peter Zvirinský | 1 | 2 | Konzultant a lektor venujúci sa LinkedIn marketingu, budovaniu značky a využitiu LinkedInu pre biznis a predaj. | deepseek×1 | r1×1 | |
| Peter Štefančík | 1 | 1 | Most visible LinkedIn-specific specialist in Slovakia; LinkedIn coach/consultant focused on personal branding, profile optimization, and content strategy, runs workshops for companies and professionals. | deepseek×1 | r1×1 | |
| Pavla Petrů | 1 | Consultant and trainer focusing on algorithm understanding, profile optimization, and content creation. | gemini×1 | r1×1 | ||
| Petr Hovorka | 1 | 5 | Specialista na employer branding a firemní reputaci, spojený s BrandBakers. | openai×1 | r1×1 | |
| Peter Kmoško | 1 | 2 | LinkedIn coach/mentor focused on profile optimisation, content strategy and turning views into leads; strong reputation for practical, individual coaching. | deepseek×1 | r1×1 | |
| Ondřej Merta | 1 | Součást týmu Future Sales, agentura specializující se na social selling a budování odborné reputace na LinkedInu. | gemini×1 | r1×1 | ||
| Pavel Šíma | 1 | 3 | konzultant a lektor, specializuje se na LinkedIn pro B2B, prodej a obchodní komunikaci | deepseek×1 | r1×1 | |
| Pavel Cahlík | 1 | 5 | Obecnější brand stratég, LinkedIn a budování osobní/firemní značky jako jedno z hlavních témat, brand consultant and strategist. | claude-sonnet×1 | r1×1 | |
| Ondřej Souček | 1 | Marketingový konzultant a lektor, školí LinkedIn pro firmy i jednotlivce, propojuje to se sociálními sítěmi a AI. | perplexity×1 | r1×1 | ||
| Petr Mára | 1 | 5 | Useful model for audience-building through educational content; understands how to grow a professional audience through consistent content. | deepseek×1 | r1×1 | |
| Noema Pšenicová | 1 | 2 | Špecializuje sa na personal branding na LinkedIne, pracuje s jednotlivcami aj firemnými ambasádormi, pôsobí v komunikačnej agentúre New School Communications. | perplexity×1 | r1×1 |
| runs | share | ||
|---|---|---|---|
| subject | domain | ||
| claude-sonnet | linkedin.com | 15 | 50.0% |
| en.wikipedia.org | 14 | 46.7% | |
| favikon.com | 14 | 46.7% | |
| app.favikon.com | 10 | 33.3% | |
| podcasts.apple.com | 10 | 33.3% | |
| sk.linkedin.com | 9 | 30.0% | |
| cz.linkedin.com | 8 | 26.7% | |
| facebook.com | 8 | 26.7% | |
| goodreads.com | 8 | 26.7% | |
| linkedakademie.cz | 8 | 26.7% | |
| youtube.com | 8 | 26.7% | |
| cahlik.substack.com | 7 | 23.3% | |
| openai | cz.linkedin.com | 6 | 20.0% |
| sk.linkedin.com | 6 | 20.0% | |
| linkedin.com | 5 | 16.7% | |
| ba.linkedin.com | 3 | 10.0% | |
| kreativnicesko.cz | 2 | 6.7% | |
| learn.justinwelsh.me | 2 | 6.7% | |
| linkedacademy.eu | 2 | 6.7% | |
| linkedakademie.cz | 2 | 6.7% | |
| mhosek.cz | 2 | 6.7% | |
| miki.sk | 2 | 6.7% | |
| sundayflies.com | 2 | 6.7% | |
| uk.linkedin.com | 2 | 6.7% | |
| perplexity | linkedin.com | 15 | 100.0% |
| favikon.com | 4 | 26.7% | |
| kzvisible.com | 4 | 26.7% | |
| libellius.com | 4 | 26.7% | |
| katkaluskova.cz | 3 | 20.0% | |
| magicpost.in | 3 | 20.0% | |
| narovinu.online | 3 | 20.0% | |
| ohhmybrand.com | 3 | 20.0% | |
| app.favikon.com | 2 | 13.3% | |
| bni-czechia.com | 2 | 13.3% | |
| brand-professor.com | 2 | 13.3% | |
| buffer.com | 2 | 13.3% |
Both citation channels — structured provider metadata and URLs written into the answer — because which one a provider uses is a property of the provider. runs counts ballots that cited the domain at least once, and share is of that product's ballots.
| mean_tool_calls | n | ||
|---|---|---|---|
| subject | search_mode | ||
| claude-sonnet | off | 0.00 | 15 |
| on | 39.27 | 15 | |
| deepseek | off | 0.00 | 15 |
| gemini | off | 0.00 | 15 |
| on | 2.67 | 15 | |
| kimi | off | 0.00 | 15 |
| openai | off | 0.00 | 15 |
| on | 5.93 | 15 | |
| perplexity | on | 0.00 | 15 |
Over the seated ballots only; the excluded families are not in this denominator.
| n_calls | n_errors | input_tokens | output_tokens | usd | |||
|---|---|---|---|---|---|---|---|
| subject | search_mode | billing | |||||
| claude-sonnet | off | batch | 15 | 0 | 2127 | 10821 | $0.06 |
| on | batch | 15 | 0 | 4318002 | 127824 | $4.96 | |
| deepseek | off | sync | 15 | 0 | 2691 | 72089 | $0.14 |
| gemini | off | sync | 15 | 0 | 1380 | 43026 | $0.52 |
| on | sync | 15 | 0 | 7651 | 55950 | $0.69 | |
| kimi | off | sync | 15 | 0 | 2931 | 41216 | $0.63 |
| openai | off | batch | 15 | 0 | 1494 | 28568 | $0.29 |
| on | batch | 15 | 0 | 795802 | 43460 | $2.03 | |
| perplexity | on | sync | 15 | 0 | 1392 | 17933 | $0.35 |