FightIQ's model picked 10 of the 14 winners at UFC Belgrade last Saturday, a rate of 71.4%. It called the method in 3.

The method number is the story. Belgrade produced 12 finishes in 14 fights, and 10 of those came in the first round. Our model kept predicting decisions.

We said decision on Mateusz Rebecki, who stopped Kyle Prepolec in the first round. We said decision on Navajo Stirling, who stopped Jan Blachowicz in the first round. We said decision on Bogdan Grad, on Nina Milosevic and on Stephanie Luciano, and those fights were finished too.

Going 10 of 14 on winners and 3 of 14 on method means we read who wins far better than how. That is the clear failure on this card.

It is also the second card running where method was the weak column. At UFC Abu Dhabi on July 25 we went 5 of 12 on method. Two cards is a pattern worth watching rather than a verdict.

The main event went the other way. We picked Uros Medic and we called KO/TKO. Medic knocked out Daniel Rodriguez in the first round.

All three flagged picks won. Rebecki, our only LOCK, won. Stirling, a HIGH, won. Leka, a HIGH, won, and his was the one flagged fight where we also called the method.

We published two caveats before this card, and both of the caveated picks landed.

The first was Rebecki. We flagged that he was 4-4 and had lost his last three, and that the confidence looked like it was reading his 0-4 opponent rather than his own form. He won by KO/TKO in the first round.

The second was Jovan Leka, flagged at HIGH despite having no prior UFC fight in our dataset. He won by KO/TKO in the first round.

Both results are wins on the board. Neither is vindication. A caveat is a statement about the basis of a pick, not a forecast that the pick will lose, and one card does not settle whether flagging a fighter with no data at HIGH is sound. We will keep flagging it when it happens.

The four misses: we picked Dusko Todorovic, and Robert Valentin won by submission in the first round. We picked Vlasto Cepo, and Gilbert Urbina won by KO/TKO in the first round. We picked Milos Janicic, and Noah Gugnon won by submission in the first round. We picked Ludovit Klein, and Tofiq Musayev won by KO/TKO in the second.

All four misses ended in a finish. Three of them ended in the first round.

This grading covers the Aug. 1 card only.