A thread throughout this Pitching by the Numbers 2014 Preseason Primer has been that we find surplus fantasy baseball value with pitching because it’s cheap.
There’s a debate within the industry on why a near universal 70/30 percent hitting/pitching split exists in auctions (which I know that so few of you play, but that doesn’t matter for this exercise). Since hitting is 50 percent of the categories, why shouldn’t we use 50 percent of our budget on it? Extending to drafts, why shouldn’t, say, every other one of your top picks be spent on a pitcher?
I’ve tested focusing on pitching as equally as I can stomach in expert leagues (which are almost always free/no money) just to see if it really is much harder to construct a winning team that way. I’m convinced it is. But I stipulate that, in this game, any strategy well executed could win. But no one is winning because they are swimming upstream looking for bargain hitters after paying top dollar top pitchers.
The are a number of assumptions made by those advocating we spend more in auction money or in draft picks on starting pitching, all of which I think are misguided (and I’ll prove it).
The first is that the fantasy community is rationally budgeting hitters/pitchers at 70/30 percent. All of us. In every league. And we’re all sticking to our budgets practically to the penny. Sure, there may be a league that goes 65/35 when a couple of owners have a pitching agenda and, conversely, 75/25 when they have a hitting one. But its a very narrow range. Let’s for a moment think of how incredible that claim is. Basically, it’s saying that fantasy baseball players have some supernatural discipline.
The next assumption is that we are discounting the value of pitchers because we know our projections are probably wrong compared to our hitting projections. So we won’t pay full value for our pitching projections. We make the projections, but we don’t really believe them so we cut their value by exactly 40 percent. Now we have the fantasy baseball community as exemplars of mathematics and rationality, too. We sure think highly of ourselves, don’t we? Again, crazy talk.
But we indeed do have this split. So, why? Is it because we have 14 hitters and 9 pitchers? That’s about 60/40? No again. As @chris_liss has said to me, “If we had 90 hitters and 10 pitchers but the categories were 50/50 would the split be 90/10?” It would not be.
Here’s the reason we have the 70/30 split, why it matters for how we value hitters and pitchers in auctions and drafts and why we not only do not gain an advantage by ignoring it but actually hurt our chances for success when we do.
The 70/30 spit is organic. Or at least it has nothing to do with budgeting or the value we place on our projections. We all basically agree on the hitter values so we are going to be pressed to pay top dollar for them by our fellow owners. This does not mean we think that hitters are more projectable or that the projections themselves are more meaningful/valuable. It simply means that if everyone agrees that Hitter X’s stats are worth $25, one of us is going to have to pay it.
With pitchers, though, there is much less agreement with projected value. Again, this does not mean pitchers are less projectable or that we are rationally discounting our projected values. It means that if I think Danny Salazar has a projection worth $20, I’m very likely not going to have to pay close to that because not enough people are going to agree with me to make me pay.
For the auction wonks out there, I will stipulate that a handful of hitters -- and to a much lesser extent pitchers -- can go for much less than most think they are worth due to owners having spent too much. But that’s balanced out by the roughly equal number of times they spend too little and thus have to overpay in the endgame.
Okay, last metaphor for why this happens. Say you needed to find surplus value with a fixed amount of money to spend and could only go to two auctions: one involved precious metals with prices that are basically fixed and the other involved collectables. There are no bargains in the precious metals auction because there is wide agreement about the value. But if you really know your comic books or baseball cards or whatever, you can make money at the collectables auction, right?
But this is a numbers column where we quantify stuff. If I’m right, it should be easy to prove. With the help of Greg Ambrosius of Stats, LLC let’s look at NFBC draft values, bringing this entire debate back into the real world of the vast majority of leagues. My theory before looking at the data is that the range of where the hitters are drafted will be much tighter than the range of where comparable starting pitchers are drafted. And I’m going to be as fair as I can be by making the hitter groups twice as large as the starting pitcher groups. That increases the variance of the hitters (we’re looking at more of them) relative to the pitchers, hurting my argument. But I want to push hard against what I think. Here’s the average differential between the highest slot anyone drafted a hitter versus the lowest slot where that same hitter was drafted in NFBC drafts through last week.
Top 20 hitters: Average high/low draft position variance, 17 slots.
Hitters 21-40: Average high/low draft position variance, 40 slots.
Hitters 41-60: Average high/low draft position variance, 60 slots.
Hitters 61-80: Average high/low draft position variance, 88 slots.
Hitters 81-100: Average high/low draft position variance, 102 slots.
Okay, I’m worried. This is more variance than I expected, to be honest. But let’s proceed with the pitchers:
Top 10 starting pitchers: Average high/low draft position variance, 33 slots.
Pitchers 11-20: Average high/low draft position variance, 69 slots.
Pitchers 21-30: Average high/low draft position variance, 119 slots.
Pitchers 31-40: Average high/low draft position variance, 111 slots.
Pitchers 41-50: Average high/low draft position variance, 131 slots.
Again, this is not total variance. This is average variance for every one of the players in each of the groups.
Add it all up and the average variance in drafts for all top 100 hitters is 51 draft slots. versus 93 slots for the top 50 starting pitchers. That basically works out to a 65/35 split. Perhaps the rest is the minor effect of having more hitter spots to fill or some small difference between the value of draft slots versus auction dollars.
More to the point, let’s focus on those top 10 pitchers, who generate slightly more agreement with us as our second and third round hitters. That’s why we have to pay for them. Everyone knows they’re not going to fall. Again, and I know this can seem like an abstraction (I don’t think it is), it’s not because they are more projectable. That’s important to know because when people assume that’s why we wait on pitchers and show us it’s wrong because they actually are projectable, they are entirely missing the point.
So, strategically, when plotting your drafts, you can gain an edge by forgetting about the low variance pitchers in the top 10 and focusing instead on the much higher variance pitchers just one group below them. You are guaranteed to get at least two of these guys way, way cheaper than you think they are worth because your league will not agree with you. It doesn’t make them wrong and you right. But it is undeniably the best way for you to shop for surplus value at auctions or in drafts.