If I asked you which stage in the first season seemed to produce results that were different from other stages, what would you answer? I’d bet that most people would go with Stage 4. You had Dallas (bad in Stages 1-3) and New York (incredible in 1-3) both going 6-4, Boston (10-0 in Stage 3) going 4-6, and the L.A. Gladiators (17-13 through three stages) going 9-1, just to name a few unusual results.

To determine just how much Stage 4 deviated from other stages – and to see how different those other stages were – I used the following approach. I multiplied a team’s wins in a stage by three and then found the difference between that number and its total wins in the other three stages. In other words, if a team had 18 total wins in Stages 1-3, an average of 6 per, we’d expect it to have 6 more wins in Stage 4. Any difference from that, in either direction, would be seen as unusual.

Here are the results for each team in each stage, along with the total difference in expected wins for each stage and the average per team. Note that the “Totals” row and column use individual win totals, while the numbers in the main section of the chart are larger, to provide a clearer set of numbers.

Indeed, Stage 4 was the biggest outlier, with teams being, on average, two wins different from their expected performance based on the previous three stages. Stage 1 was the least atypical stage, with teams doing about the same there as they did during the rest of the season.

Looking at this metric on a team-by-team basis also reveals some interesting facts. The Gladiators were the most atypical team in Stage 1, when they went 4-6, going 7-3, 6-4, and 9-1 in the other three stages. Boston was all over the map, with their 10-0 record in Stage 3 seeming like a major outlier, with 6, 6, and 4 wins in the other three stages. Shanghai was, unsurprisingly, the most consistent team throughout the season, but Philadelphia (6, 7, 5, and 6 wins) was pretty close.

Most analysts would probably point to the “Brigitte meta” in Stage 4 as being the reason for the topsy-turvy records in that stage. In fact, in general, the Overwatch League season is an unusual beast, as compared to other sports leagues. There were patch changes occurring throughout the season with every stage – remember Mercy and her pre-nerf Valkyrie in Stage 1? – which make the league vulnerable to significant swings in play styles and standings shakeups with each stage.

That said, every sports league has teams that fluctuate during the season. An NFL team might start 0-8 and then go 4-4 in the second half. A baseball team could be riding high early but then suffer some injuries and falter. In every sports league, players come and go, whether due to injuries (uncommon in OWL), off-field issues (xQc, Undead, and Dreamkazper come to mind) or new acquisitions (Fissure in L.A. or most of the Shock‘s second-half lineup). And that’s not even taking scheduling and random luck into play.

There would be some inconsistency in per-stage records, even if Overwatch didn’t, essentially, change its rules ever quarter-season, something that’s all but unheard of in most sports leagues. In the coming weeks, I’m going to try and dive into that and see if I can’t separate how much of the differences in records had to do with OWL’s unique format versus other factors.

Whatever I find out, I think we can all agree on one thing: It was a blast watching Mickie play Brigitte, wasn’t it?