BONUS SUNDAY FUNDAY STREAM - The economy of roster spots in MLB DFS
Epiphanies about “positional scarcity” (is that the right phrase?) in DFS baseball.
I’ve come to the realization too that … while again maximizing points and maximizing value, in general, are always core concepts …
I’ve been thinking more in GPPs lately (since I’m playing less and thinking about it more) … thinking of each position spot as more of a “budget” as opposed to the salary … I mean obviously the salaries have to fit … but I think good GPP lineups will be built around … well basically having the highest scoring player at each position.
OBVIOUSLY. But hear me out.
So not just being the highest scoring player, but to place high in GPPs we need to select players that end up giving us a great SEPARATION in production at each available position spot.
It sounds kind of obvious but I think … if my reasoning is sound here … there are some things to take away from here …
I guess to cut to some of the chases … there’s not much reason to be contrarian at outfield. Outfield is so deep and everybody gets to pick 3 of them … so because of that … the way you get SEPARATION from the rest of the pack at the outfield position is by getting an INSANE performance. So if you pick three outfielders and they’re all 5% owned and 2 out of 3 of them hit home runs – that feels like a pretty impressive feat. But every day there are …. a dozen or more OFs who hit home runs on a full slate? Anyway, several to a lot, of outfielders, who homer. And many of them are very-high-owned given the fact that we get to pick 3 of them. So really at the end of the day in a GPP there’s going to be a huge chunk of people who got 2 HRs from their 3 OFs and it doesn’t matter if your guys were 5% owned and their guys were stanton and harper and trout … you didn’t create any separation by getting cute here.
So what we see here is the main benefit we can get at outfield is, perhaps, saving money. While it is risky to bypass those high-owned, high-HR guys in good situations … it actually might be the best play to balance your OF with that type of “harper/stanton/trout” type $5k player with your rasmus/carter(1b only now I think)/raburn???? whoever else you think might be a high-HR, low-salary option … but it’s about getting home runs for the salary from your OF spots.
We’ll see the same thing with pitchers. Again, on most big slates, there will be several high-scoring pitchers, it will be very rare that you get the one pitcher who throws a CGSO and has 2x as many points as any other pitcher. That said, we don’t want to get behind the 8-ball at pitcher. So since the possibility for legitimate SEPARATION on a large slate MLB DFS GPP is low at the SP spot – we hope to get the guy with 15+FD points or 30+ DK points but on a lot of big slates there will be 3 or 4 pitchers who hit those marks and most of the entires will have one of those guys. So again our goal at pitcher … we can’t afford to not get points here, so when in doubt I will pay for pitching even in GPP, but our one chance to get an edge is to get one of the guys who is going to be in that 15+ FD, 30+ DK (just ballparking numbers) category, but we get him for cheaper than most of our opponents. This is a tricky one but there is a lot of benefit if … you have say, a Sale, Scherzer, and Hamels putting up 30 points for $10k, and we can get a Lance McCullers to pop for 28 points for $7k. Again made up examples for the most part, but you get the idea. I think trying to predict CGSO’s or 16+ strikeout games is a losing proposition. Our best play at pitcher is to not be contrarian, but keeping in mind that our one legitimate chance for an edge is trying to get a value guy in a good spot to match the production of the chalk pitchers.
I’m tempted at first thought to kind of put first basement in with OFs and Pitchers as … it being a very deep position, everybody’s going to have a 1B who they plan on contributing a homer when they submit their roster for an MLB slate on most any day. Obviously not everybody is going to be right, but the position is so deep with elite power hitters that … enough people WILL be getting significant production from their 1B on a given day, that I think it is agreeable to play this position the same way as OF and SP – we basically don’t want to NOT get points here, so we will pay up a little bit and lean more towards the chalk, because a good number of entries are going to have SPs, 1Bs, and OFs who are earning them significant points. We basically need a HR+ game from our 1B to keep pace – again we will see that there is the somewhat-legitimate possibility of getting separation here obviously if we get the guy with the 3HR day or something – just like with OF – but just like with trying to predict a CGSO or a 16+ strikeout game … I think we’re better off not trying to predict those, and just rely on the fact that if we continually play good players in high-leverage situations, the DFS gods will smile on us with those “outlier” days more often than a weaker player. So again the one place we can get an edge at 1B is by saving salary – if we can reach down and get an Adam Lind for $2700 fanduel at home against a RHP like he was recently … and we can get 6ish fanduel points out of him and keep pace with all of the people who are owning encarnaction and goldschmidt (as examples of generic high-owned, high-salary 1b’s) … then that is a definite Win, but again we need to pick our spots here and not trot out random 2500 1B’s just hoping they hit a HR – because just like with OF and SP … we’re not going to WIN gpp’s by getting normal high-end production from these positions. But we will very likely be out of most GPP races if we throw a crappy SP up there and he scores 8 DK points and 40% of the field has kershaw and he scores 40 – we’re pretty much dead. So at these three positions we can lose ourselves GPPs, but even if we roll the dice with cheap guys, and they do well – all we have is extra salary left over, and that doesn’t win us GPPs either. How can we turn that extra salary into GPP-winning points?
2B, SS, and catcher.
So these positions have some legitimate scarcity to them. 3B is definitely kind of the true inbetween spot between … if we have 7 total positions on the spectrum … 3b is the midpoint, with three squarely on one end of the spectrum and three on the other end. There is SOME scarcity there on some days, it is definitely not as deep as 1B .. but there’s enough quality power hitters there that it’s unfair to compare it to the others.
2B, SS, and catcher … things get interesting though. The fact that we only get to pick one player at these positions, and on many days there will legitimately be “have’s” and “have-nots” after all the scores are tallied … we have a chance to get legitimate SEPARATION from the field at these positions. There are a few ways to do this …
This is a spot where, say at shortstop especially, a lot of times we will have one Tulowitzski or Hanley, maybe a second tier of guys like Peralta? The chalk, whoever it is, on a given slate – the list goes 1-3 players deep usually of legitimate “chalky” options. So reaching down for a $2200 shortstop here … you are only betting against a couple of MLB hitters doing bad on a given day, which, as we all know – will happen quite often, no matter how good the hitter is or how good the matchup is – failure is always more common than success. So we have a fighting shot going off-the-reservation here, whereas doing that in the outfield, we are trying to rely on 5, 10, 15, 20 different excellent players not doing better than our guy. Bad idea.
We see this separation or “scarcity” maybe to different degrees at 2B and catcher as well, obviously it fluctuates from slate-to-slate depending on exactly what teams are playing against what pitchers and in what stadiums. But in general the pool is significantly thinner at these three positions.
So like I said it is a good play to reach down in salary here – but the money is irrelevant. Just because of the fact that if you can grab a 5% owned guy at SS you actually have a shot in hell at him being the highest scoring SS of the day and the 2nd and 3rd highest guys might be several points behind – SEPARATION!
But what I think might be the even better play … to build the optimal GPP lineup – is to go after this scenario: figure out a way to grab a McCullers, or a Lind, or a C Carter or somebody who is generally a high-variance, high-reward player … and try to steal 7IP 7K’s and a W or steal a HR at one of those spots … keep pace against the rest of the field in those positions, and use that extra salary at SS and 2B and C … … … people have a tendency … to spend the money at SP, spend it at 1B, spend it in the OF … and we can use this to our advantage. Those positions are so deep that we can get value there and steal a HR and keep pace. Then take the extra money, and whereas I think your generic DFS player LOVES to grab a jose ramirez or that iglesias guy or whatever – I call them the daily “surprise 2-hitter” … these types of middle infielders get popped into the 2-spot like once a week in high-powered offenses and get a platoon advantage ….. and they are very strong plays, and usually min salary, and everybody knows it. So that’s the recipe for most people and it felt right to me for a long time – grab up that cheap high-leverage SS and 2B and make sure we have stanton and donaldson (or whoever) in our lineups because they’re in insanely good spots and insanely good players and … we HAVE to have them. In fact they are probably easily the highest-likelihood players to hit a HR on a given day. And yeah donaldson might go for the 35-point-DK-blastoff once in a blue more more often than luis valbuena will … but at the end of the day it’s really, the two most likely outcomes that we care about are … donaldson and valbuena both finish with between -1 and +1 fanduel points … or 0-6 DK points … or they both hit a homerun and maybe get some other small points or maybe not – but they end up between 6-8 fanduel points or say 14-20 DK points. OR – one of them does, and one of them doesn’t. Those are the “likely” scenarios that are meaningful to us. So, yeah, donaldson will hit more home runs and score more points per game than valbuena. But if we can get valbuena at home against a RHP … … … he’s got APPROXIMATELY as good a chance as anyone at hitting a home run, based on the metrics I use (if you disagree then insert player names that are more appropriate – I know we all value everyone differently) … so there’s a lot to be gained there.
So knowing that home runs probably vary, for good power hitters, from between like 1-in-6 to 1-in-3 chance per game (with the one in three being ONLY the most extreme power hitters in the most extreme high-leverage situations due to hitters park, extremely poor pitching, etc.) … there is definitely a difference between an adam laroche and a paul goldschmidt. But if LaRoche or Lind is 2700 at home and Goldie is 5400 on the road .. is Goldie really twice as likely to hit a home run and therefore score 6+ DK points than Lind is???? Maybe not the perfect example … and a lot depends on situations … and like I said insert different player names, but I think all of us know of that one cheapie 1B that we like to start when they have a platoon advantage in a good park – everybody’s got one, we all just have different ones :)
So then if we can swing those – keeping in mind it’s a fine line to dance around trying to decide when to “fade” the elite hitters and when not to – something that very few of us have mastered, for sure, and I definitely haven’t – …. if we can swing any money-savers and keep pace at SP, 1B, or OF … that is the first step … but I like to use that extra money and pay UP for the middle infielders. Grab up an Altuve, Kipnis, or Tulowitzski … maybe even against an above-average, but not an ace, pitcher. It is the definite trend to see huge ownership at SS and 2B for the super cheap, good-situation guys … and for good reason. But I’ve played those lineups enough to know – often times they will do well – your cheapie infielders drop a bunch of points on ya and a bunch of your studs homer and you were able to afford the ace pitcher playing the phillies and – yay – you scored 150 DK points! And guess what … you probably just won $6 in the moonshot. Because … you didn’t get any SEPARATION! At least … that has been my experience. It’s exciting but when you do well you still won’t be way ahead of anybody. On those days I think you win the GPP by lucking out and getting the 40 point DK hitter or the guy who hit 3 homers or just ACING every position top to bottom on your roster. It becomes more of a lottery. But we will have many many more days where the chalk fails – and those are the days that I play GPP every day hoping to get.
But the point of this is not at all to simply say “being contrarian pays off in GPPs” … it’s a matter of WHERE to be contrarian … and being aware of the fact that sometimes paying $5400 for Troy Tulowitzki on the road instead of playing Jose Ramirez batting 2nd with a platoon advantage for the indians at 2200 is one of the best contrarian plays we can have.
Before I end I want to say that yes, I am aware that there are more ways to score points in DFS than by hitting home runs. So a lot of this analysis can be extended to total points … but honestly home runs are extremely important and represent the one true discrete event that we can say … “if it happens, we did good at this position” … so I like to use it heavily, and use it a lot when making examples. You could say the same for the other points-scoring stats, but in general, HRs are the most important, so I refer to them most frequently.
So just a final note … I guess if I had to sum this up … and wanted to sound scholarly … I would talk about it as “an economics of roster spots” kind of approach … we have to look at each of our 9 or 10 (depending on the site) roster spots and say “are enough of these spots occupied with players who give me a reasonable chance to SEPARATE from the field at this position?” … everybody only gets to pick one shortstop! So that is huge. Everybody gets one first basemen, but it’s very likely that a good chunk of teams will get production there, as we have explained. So if we admit (which I am willing to do) that our chances for separation at 1B, OF, and SP are limited only to extremely improbabable and unpredictable uber-performances … then we really need to maximize our separation at 2B, SS, and C, remembering that it’s all a crap shoot and the only way we’ll ever see results that are indicitave of our lineup strength are by playing for a very, very long time. It’s a tricky game but I think by going through this thought exercise I’ve identified a few ways that we might be able to do this – keeping in mind that, as always, every slate is different.