Revisiting Ottoneu Arbitration Strategies: Part II

In Part I, I went over the data I used and talked about two strategies – putting on arbitration allocations on players to try to force them into the auction and allocating to pre-breakout potential stars. Today, we’ll look at a couple more topics.

Hitters not Pitchers
“Don’t pay for pitching” has been a staple fantasy baseball strategy at least since the glory days of the Ron Shandler LIMA plan. It’s not the only strategy, as you’ll also hear people talk about the “pocket aces” draft strategy which is basically the opposite.

But it remains a popular approach and it influences arbitration, too. If you don’t believe it makes sense to spend on pitching – because it is volatile, because you can find it at the draft, because injuries are too common, or any other reason – why would you allocate arbitration dollars to SP?

On my article of SP allocation targets last week, reader CC AFC expressed this idea well in a comment:

My goal is pretty much to never allocate to pitchers. I want my arb dollars to stick with the other [managers] as long as possible. The pitchers will break at some point, at which point they will be released and auctioned again during the season or the next year, and my arb is gone… [Probably] 95% of my arb is going to hitters.”

I generally agree, but there is reason to believe that thinking has gone too far. There are two main data points driving my thinking on this.

First, from the league one 2021 arbitration data, of the 69 players who received at least $1 in arbitration, 24 were pitchers and 46 were hitters (and yes, I know that 24+46=70, but one of the 69 was Shohei Ohtani, so there you go). That makes sense – first of all, teams roster more hitters than pitchers, second CC AFC isn’t alone in preferring to allocate to hitters.

In that article, I defined “fully effective” allocations as those that not only stuck with the player through 2022 but are going to stick into 2023, as well. I defined “effective so far” as allocations that stuck with the player through 2022, but won’t stick into 2023.

Those bottom two rows are subject to my opinion – I had to make a judgment call as to whether or not the player will be kept this off-season but even if you assume I was 20% too aggressive in assuming hitters will be cut and 20% too aggressive in assuming pitchers will be kept, pitchers are still more likely to show up as “fully effective” (54.2% vs. 47.8%).

The pattern, by the way, is the same if you look at it by dollar instead of by allocation:

The other data point comes from Justin Vibber, of Surplus Calculator fame, who tells me that surplus (player value above player cost) tends to be evenly split across hitters and pitchers at this time of year. Given that the arbitration process serves to reduce surplus, if surplus is evenly split amongst hitters and pitchers it would stand to reason that arbitration allocations should be evenly split, as well. Now, that isn’t entirely true – because pitcher performance can be noisier and riskier, some risk-aversion when it comes to allocating to pitchers is warranted.

But it seems like, at least in league one, we have swung the pendulum too far. If we took the 8-10 riskiest hitters we allocated to, or the $15-$25 in riskiest hitter allocations, and moved that over to pitchers, we would likely end up with more allocations and more dollars in the “effective so far” or “fully effective” category.

Instead of putting allocations on an already-expensive Mookie Betts (who was cut before the auction) or a risky Alex Kirilloff (who never earned out his salary), we could have targeted more dollars on Alek Manoah or put money on Jordan Montgomery. Both of those pitchers still have surplus, per the Surplus Calculator.

Those specific examples benefit from hindsight and there is certainly more risk up-front, but the data suggests that allocations could be more evenly split between hitters and pitchers. I will definitely be less pitcher-averse in my allocations this year.

Go it Alone or Pile On
I typically try to avoid concentrating arbitration allocations on just one player. If multiple managers have already allocated to a player on another team, I like to try to find another target. Not always – that $10 Julio Rodriguez is going to get a lot of allocations and I won’t hold back. But in general, I think that diversifying the allocations spreads the money around another manager’s team, increasing the likelihood that at least some of the arbitration allocations are still on their roster a year down the line.

The rationale here is that if everyone piles on a single player, there is risk that all the dollars are washed away in one move. If we put $33 on Rodriguez and his sophomore campaign is a disaster, he is an easy cut in a year, and that manager has cleared all $33 of 2022 allocations. If instead we put $15 on him and the other $18 on other players, some of that money sticks around.

However, the data suggests I should think twice before taking that approach. This table compares the effectiveness of allocations placed on players that received $5 or more in allocations vs. those that received allocations from just one team.

The results aren’t drastic, but there is at least a small “wisdom of the crowds” effect here. It gets even more clear when you consider that three of the four players hit with “ineffective” allocations were targeted by just one or two managers, and none of them were targeted with more than $4.

This data does not directly contradict my concern – the players who were hit with $5 or more account for $228 of the $300 total dollars as compared to $27 for the players who received just one allocation, so the risk is in fact higher. The “effective so far” line on that table reflects a lot more risk in the $5+ group.

But it also shows that when you go it alone, you are taking on more risk for your specific allocation.

I don’t think this changes things for me strategically, but tactically I will stop and think twice before moving dollars off a highly-targeted player for an ignored player. If I am the only manager allocating to a player, I want to stop and make sure of a few things:

1. Am I missing some news? The allocation to Buster Posey last year is a highly visible example, but trade rumors, injury concerns, etc., can all factor into making a player a less attractive arbitration target and if every other manager is avoiding that player, maybe there is a reason.
2. Am I the high-value on this player? If the player has a $5 salary and no one else thinks they are worth $6, maybe I should try to acquire the player in a trade. Or maybe I need to rethink my valuation. It’s possible the player has plenty of surplus value but there are just more attractive options on the roster. $6 Taylor Ward probably deserves some allocations, but he is on the same roster as that $10 Julio Rodriguez – Ward may be stuck at $6 because no one thinks he is worth $7, or because everyone just prefers to allocate to Rodriguez.
3. Am I being different just for difference’s sake or is there a real reason? Using that Ward vs. Rodriguez case, if Rodriguez has been pushed to $25, my decision to push him to $26 vs. Ward to $7 is different than if he is pushed to $40 and I am debating pushing him to $41 vs. pushing Ward to $7. The risk that Rodriguez turns into a solid $30 player who is a future cut at $41 is much higher than him becoming a $20 player who is a future cut at $26. I don’t want to zig just because everyone zagged – I want to zig because zigging is the smarter move.