Last time, I took a look at how much Overwatch League teams’ records changed during each stage. My ultimate goal was to try and figure out if OWL is any different from normal sports leagues in this respect. In other words, how do OWL teams’ win totals by stage compare to other leagues’ wins by quarter-seasons?

Fortunately, the three major sports leagues that use straight win-loss totals – I decided not to mess with the NHL – all have seasons that are (nearly) divisible by four. For this study, I took the first 80 games of the most recent NBA season, the first 160 of the last MLB season, and all 16 games of the last full NFL season.

I split them up by quarters, just as the OWL is divided into stages and then “normalized” each quarter as if it had the same number of games (10) as the OWL. So the NFL (4 games per quarter) got multiplied by 2.5, NBA (20) was divided by 2, and MLB (40) was divided by 4. I then performed the same computations on those quarterly win totals as I did with OWL teams.

As a reminder, I take the difference between a team’s win totals in each quarter and compare it to the sum, divided by three, of its wins in the other three quarters. The greater that result, the more that team was “inconsistent,” with fluctuating win totals throughout the season.

Of course, in the non-OWL leagues, the rules don’t change during the season. If OWL is unique in that respect, we should expect to see bigger numbers in the OWL portion of the chart.

However, there’s another factor at play; leagues that have longer seasons, and thus more games per quarter, should have less variance in team wins per quarter, and the opposite should be true for leagues with shorter seasons. For example, nobody is surprised if an NFL team is 1-3 in its first four games and 3-1 in its next four. A baseball team going 10-30 over its first 40 games and 30-10 over the next 40, though, would be highly unusual.

Here’s one more factor that I thought would affect things: In OWL, only six players are playing at once and teams tend to play the same players a large portion of the time. The NBA is similar, with just five players on the court at a time and the top players out there around 75% of playing time at their positions. Compare that to MLB, where even a durable starting pitcher only gets about 1/7 of his team’s innings and teams use nine batters. The NFL is even more split; between offense, defense, and special teams, nobody is even on the field for half of all plays, and there are 11 men per side.

What all this means is that, in OWL and NBA, individual players should have a greater impact than in other sports (with the possible exception of the quarterback in football). Therefore, you’d expect more volatility in win totals for those two leagues, as a single player performing poorly or being unavailable (usually due to injury) can have a much greater effect than on a team that uses a dozen or more players in every game.

All right, now let’s get to the chart:

The “Team Average” column represents how much an individual team’s win total tends to vary over each quarter-season. The NFL is #1 in this regards, followed by OWL, NBA, and MLB. They go exactly by season/quarter-season length, which seems to indicate that that’s the determining factor in how much win totals fluctuate for any team. Basically, it’s all about the sample size. Fewer games equals more chances for volatility in win totals.

Look at the last column, though. That’s the variance of those four quarters taken together. If the previous column showed us how much individual teams’ win totals varied during each quarter-season, this one tells us how “spread out” those quarterly differences are throughout the season. The OWL has a much higher variance than the other three leagues, even the shorter-season NFL.

You can tell how the leagues compare just by looking at individual totals. For OWL, they range from 1.08 to 2.03, a range of 0.95. For NBA, that range is 0.32, for NFL it’s 0.31, and for MLB it’s 0.14. The range at which OWL teams’ records tend to change by stage is much greater than it is in the other leagues. Even if I set Stage 4 to 1.42, that range would still be higher than the other leagues (0.34) and the variance would be slightly higher than the NBA’s (0.02797 to 0.0.2785).

This seems to confirm that the OWL is the most wildly varying league in terms of the differences in team wins per quarter-season – with some caveats. We’ve only got one season to go on, and I only pulled data from one season each of the other three leagues (it takes a good bit of time to compile!). Still, for now, I think the results confirm what most people have thought.