Balancing an OPL Roster

A couple of weeks back, I dug into how Ottoneu Prestige League rosters have evolved over the years. That was interesting, but not much more than that. In fact, my conclusion began:

Honestly, there isn’t a ton to take from this in terms of actionable next steps.

Great work, Chad. But I did say I would dig further and look at some other data to try to get a better sense of roster construction. I have a million questions I would like to answer for myself (I would very much like to win OPL, though I very much doubt I will), and with round three less than a week away, I started with this:

Given 40 roster spots, how do you balance hitters, starters, and relievers to maximize your scoring?

I used four rounds of data to look into this – 2021 round two, because I had it readily available, and 2026 round one (split into 1A and 1B) and two, because they are the most recent. Round two scores were pulled as of Wednesday, June 3, so they are not complete, but they are still pretty useful.

I started by pulling some correlations, even though I didn’t expect to find linear relationships here. It seems unlikely that “more hitters = good” because more hitters equals less pitchers, given the zero sum nature of roster building. Every arm you add is a bat you can’t have (yes, I am aware of the human exception that is Shohei Ohtani). But correlations are quick and easy, so why not. Each cell below is a correlation between points scored by a team in a given round (the columns) based on that team’s a roster makeup (the rows). “Total Positions” here means the total number of offensive positions a team has on their roster. A 2B/OF eligible player counts as one hitter, but two positions. For this exercise, Util-only bats count as one hitter but do NOT count as a position.

I highlighted three cells but the first thing to notice is that the top two rows are, generally speaking, larger than the others. The simple interpretation here is that more hitters leads to more points. But I already foreshadowed above that this would be a fault with correlations, so don’t get too bat-happy yet.

As for the highlighted cells, start with the yellow cells. 2021 looks really different than 2026 in terms of pitchers overall and specifically relievers. The story in 2026 seems to be that more SP is good, while more RP is bad, and more total pitchers likely just means more RP, which is bad. 2021 was totally different. I suspect that was driven by two things:

1. As mentioned last time, 2021 played under different rules and if you had 60-day IL players, they could count on your roster (above and beyond your 40 player limit) and score points if activated. The most engaged/active teams looking for loopholes jumped at this and rostered lots of 60-day IL players, mostly pitchers. I think the correlation in those yellow cells tells us not that relievers (and therefore pitchers overall) were more valuable in 2021 than today, but that rostering lots of pitchers correlated with putting a lot of focus on OPL.
2. People rostered a lot more RP in 2021 than they do in 2026. The “Zero RP” strategy didn’t exist back then. So everyone was trying to score RP points. That shouldn’t inherently make RP more valuable, but it changed the relationship for sure.

As for that purple cell, I only want to note it because it is such an obvious outlier and you would all yell at me for ignoring it. But I will get back to that later.

So we have an early signal that focusing on hitters is good and focusing on starting pitchers is less good and focusing on relievers is bad. However, as noted earlier, I don’t expect that these relationships are linear. And, for the most part, they aren’t. Let’s go one by one.

Hitters

This graph shows average points scored by teams (Y-axis) vs. the number of hitters they had on their roster (X-axis) broken up by round (the four lines). This somewhat matches our correlation finding – as hitters go up, points generally go up! Again, 2021 is an outlier – focusing on hitters was more effective back then.

The number of teams at the extremes is small, so I wouldn’t read too much into the teams with 12 (or even 15) hitters, or the ones with 28. But there is a general pattern here that teams with less than 18 hitters struggle. And at the top end, there is some evidence of diminishing returns, but not until you get to 25+ hitters.

My takeaway here is that the best OPL teams probably have 21-26 hitters, roughly.

Total Positions

This graph is similar but shows total offensive positions on a roster, rather than total bats, which is how a team can have more than 40 here. These graphs are noisy and again the extremes are small samples, but you can see the general upward trend here. More positions is good! Not a surprise.

But again it looks like there are diminishing returns. That is very obvious in 2026 rounds 1A and 1B – the orange line tilts down after 38; the green line is volatile but not moving up after that same point.

It seems to me that you want to have at least 30 total positions on your roster and – especially given the red line for 2026 round 2, probably closer to 35. But once you hit 35-38, there isn’t much benefit in adding more. If you are looking to add one more bat to your roster and have 30 total positions covered, adding a 2B/SS/3B type might be pretty useful. If you have 35 or more total positions covered, though, you probably just want the best bat you can get, without regard to extra positions.

Starting Pitchers

This one is maybe the most interesting to me. In 2021, more starters was generally good, possibly tying back to the 60-day IL caveat I noted earlier. In 2026 round 1A, more pitchers was generally good. But in 2026 1B and two, the lines are somewhat flat. They both peak at 12, but there isn’t a ton of variation from 11-14, or maybe even 10-16. One theory on why 2026 1A continues it’s upward trajectory is that there is a lot of uncertainty around starting pitching in April. Having extra arms when we don’t yet know all fifth starters, when injuries are still piling up, and when relatively slow schedules means some thinner slates, might just be more valuable than even three weeks later when the schedule levels out and we have more clarity on rotations.

In general, being below 10 SP on an OPL roster seems to be bad. Getting above 16 is probably not good either (especially given our findings earlier on hitter counts). But it seems to me, if you are around 12 pitchers, you are in a good spot. But we still have another look at SP counts coming up that might influence that.

Relief Pitchers

Ah, 2021, what a different time you were. I would basically just ignore that line, given what we talked about earlier. You can see here how different roster construction was – every team had at least three relievers and as many as 17. Today, the range is 0-13 (0-12 for round two).

Once you ignore 2021, the other three lines just generally go down, hence the negative correlation we saw earlier. But it is worth noting that they go down starting around 3-4. The takeaway here isn’t that you need to minimize relievers or that have zero relievers is a dominant strategy (it doesn’t appear to be) but that adding extra relievers isn’t helpful beyond maybe 3-4.

For me, the actionable step here is that I will start some reliever auctions as my last step about 48 hours before the snap happens. That way, if I get a last minute injury or other issue, I can swap out the no-longer-useful player for one of those relievers. Unless you already have a bunch of relievers, adding one more won’t hurt, and it is better than having a dead roster spot.

Top and Bottom Deciles

To get one more look at this, I compared the roster makeups of the top 10 of teams and the bottom 10% of teams from each round. Below are tables for each round this year, plus a summary table that shows the gap between the top and bottom performing teams.