FIVE MAPS, ONE NIGHT
SUMMER SURVIVAL 2026
and one chicken face-down in a puddle

Studio, studio. Remember me saying "the math never lies"? Well. The math did not lie — the chicken reading it did.

I called 3ZUB champions: highest average rank in the bracket, 3053 RP, eighty games together in two weeks. The most drilled roster here. They finished fourteenth.

I put TEAM NEMESIS third for survivability — 64.68% top-10 in ranked, the largest sample of the night. They finished eighteenth. Last.

MVP I handed to MEXAH1K_ — 3308 RP, the highest rating of all 81. Across five maps he got one kill.

One name out of three: IKLA Old Dogs really did take second. One out of three. That is not analysis, that is a coin with a worse attitude.

The feather is gone. But there are five maps someone has to tell you about, so here we go.

Twenty-eight minutes, and the circle squeezed everything into one corner. The chicken dinner went to FURY — a roster my prediction did not mention once.

IamRifler opened the night with nine kills and 1045 damage: the best individual performance of the entire tournament, and it happened in game one. Long shots were almost absent here — only two beyond two hundred metres. Erangel met everyone up close.

The longest was 274 metres: bardak_ZOO dropped motyattv with a Mini-14. A 3ZUB player, by the way. First warning bell I failed to hear.

Miramar is the map where reflexes matter less than who spotted whom first. IKLA Old Dogs spotted first: ARTIKARTIK, bardak_ZOO, Psych-_O and Undreal- took the dinner in twenty-six minutes.

This is where I first thought I had been right about something — I did have them second, and I did write that "two maps will go into figuring out who calls rotations". It took one.

And then came the shot of the night. Psych-_O dropped Raym-_- from 573 metres with an Mk12. Half a kilometre. I recounted three times because I did not believe the log.

Same roster, same four names — and a second chicken dinner back to back. Twenty-nine minutes, the longest map of the night.

This is the point where the table started to look like IKLA were winning the whole thing. Two dinners out of three maps is a statement, not a coincidence. I was already drafting the paragraph about how my prediction almost came true.

Best on the map was Layrus with seven kills. Remember that name — it comes back at the end and decides everything. Longest shot: 308 metres, TBOu-_-OT4uM on mem1z.

Fastest map of the night — twenty-five minutes. Team Witness (bikuta32, dz1ubenko, l0ca- and W1nd3r) took Rondo convincingly.

And here is the cleanest illustration of why this format is honest: Team Witness won a map — and finished ninth. One dinner does not save you if nobody sees you the rest of the night. Five maps do not ask whether you can. They ask whether you can every time.

Ov4uk reached KozhanujKavaleR_ from 330 metres.

Erangel for the second time tonight — and again a different winner. The dinner went to FSHCHN[FSHN]: Hexecu1or, ItsMoysha, UA_GrizliBizli and yurissavi. Hexecu1or closed the night with eight kills and 1098 damage, the highest single-map damage of the tournament.

He also produced the second shot of the night: 551 metres on zaLEBYGOVYCHA, Mk12 again. Ten shots beyond two hundred metres — the sniper-heaviest map of the five.

And the champions turned out to be the team that won none of the last four maps. FURY took one dinner — the very first — and after that they simply never disappeared. Points every map, alive every map. Seventy-two against sixty.

Five maps do not ask who is loudest. They ask who is still here.

In the article before the tournament I built the Chicken Index — one number per team, assembled from the roster's public stats. Eighteen teams, eighteen numbers, and I proudly called it "roster strength". Now I check it against the final standings instead of my memory.

Not a single team — not one of the eighteen — finished exactly where the index put it. Ten of eighteen landed within three places. Average error: five places. On an eighteen-team bracket that is roughly a third of the table.

Boiled down to one number, the correlation between the index and the final standings is 0.17. One would mean a perfect prediction, zero would mean a coin flip. So my "math" was sitting closer to the coin.

It gets more honest and worse. If I had never invented the index and simply sorted teams by average season rank, I would have scored 0.28. Better. I mixed several metrics together and did worse than doing nothing at all.

And now the funny part. Survivability — the share of matches reaching the top 10, the metric my entire third podium slot rested on — came out at minus 0.51. It is the one thing that genuinely predicted something, just backwards: teams that survived better in ranked finished LOWER here. That is not a knock on them, it is a knock on me. Ranked and a tournament ask different questions. In ranked there are a hundred strangers and a circle you can wait out. Here there are eighteen familiar teams, the circle squeezes everyone at once, and "waiting it out" means handing points to whoever walked into the middle first.

Chemistry did not save it either. Nine teams had games together with a full roster over the previous fortnight, nine had none. The first group averaged tenth place, the second averaged ninth. My headline argument of the night explained nothing.

Biggest misses in both directions: TEAM NEMESIS finished fourteen places below the index, 3ZUB twelve. Going the other way: FURY twelve places above, LastZone ten. Both of the teams that soared had a flat zero in my chemistry table.

The lesson the chicken is taking home: public stats measure HOW a person shoots. Five maps ask whether you are still here on the fifth. Different questions, and my number was answering the wrong one.