All the sabermetrically-minded fantasy players out there frequently visit the FanGraphs pages. I do many many time a day. It has proven very effective for me to track the underlying stats on players to help identify potential breakouts, when a player is hitting the ball hard but is having bad luck, or substantiating the real stats a player is putting up. Over the years I’ve found my self most frequently looking at hitter’s contact rates, hard hit rates, outside swing rates, BABIP, and wRC+. All of these can tell you something different about a hitter. Contact rates indicate what a player’s batting average should be. Outside swing rates can tell you if a player is hacking it up there or is showing some plate discipline. BABIP tells you what’s going on with the ole’ luck dragon.
Today I thought it might be fun and informative to run some quick numbers to see which of the many underlying stats might correlate best to fantasy production. There’s never any one stat that is going to tell you the whole story but it might at least be valuable to know what is meaningless to fantasy value.
The first step is to set a value on fantasy production. So I made a crude way of measuring it by assigning a value to all the standard fantasy categories (R-HR-RBI-SB-AVG). It’s not complicated but basically each category can be worth somewhere between 0-5 points. A zero value fantasy player would have roughly the line of 60 runs, 5 home runs, 60 RBIs, 0 stolen bases, and a .250 batting average. That’s roughly a bit below replacement level at each of those categories. For a player to to get 5 points in a category they’d put up 110 runs, 35 home runs, 110 RBIs, 35 stolen bases, or a .310 batting average. So if a player had a 25 point season they’d basically be amazing. It’s quick and dirty, but for a large sample size it will be accurate enough for our purposes.
Then I collected data from all player-seasons over 450 plate appearances from 2014-2016. FanGraphs is very helpful with this and it is super easy to download spreadsheets with any and all the stat categories you need. Once I had my data assembled I ran plots and trendlines for 17 different categories against this new fantasy value stat.
| Category | R2 Value |
| BB% | 0.071 |
| K% | 0.012 |
| BABIP | 0.142 |
| OBP | 0.444 |
| BB/K | 0.088 |
| ISO | 0.372 |
| GB/FB | 0.018 |
| LD% | 0.003 |
| FB% | 0.020 |
| HR/FB | 0.252 |
| WAR | 0.560 |
| O-Swing% | 0.000 |
| Contact % | 0.000 |
| SwStr% | 0.000 |
| Pull% | 0.008 |
| Hard% | 0.062 |
| Soft% | 0.192 |
So I just sort of did all the mildly important categories I could think of because I didn’t really know what I was looking for. The top correlative stats to total fantasy value (in standard category leagues) appear to be OBP, ISO, and WAR. WAR isn’t a surprise as this is already an all-encompassing stat, so it makes sense that if a player is valuable in real life, they are most likely valuable in real life as well. OBP stands out to me as something new to me. I generally look at BB% instead but I may have to start making OBP my usual tracking stat. The concept is roughly the same but it would be most valuable to simply look at a players full OBP instead of average and BB% separately. ISO is another one that I’m very much aware of but I typically don’t look at. Personally I look at a player’s HR’s and doubles but this tells me I should just cut to the chase and track ISO instead, is will cut out the clutter.
On the flip side the results clearly point to O-Swing%, Contact%, and SwStr% to having zero correlation to a players total fantasy value. I must admit I did not expect this and I’m not entirely sure what to make of it just yet. We know these categories do point to a player’s plate discipline and are strong indicators to what kind of batting average a player could sustain, but perhaps there are enough Martin Prado type players that make these indicators unhelpful on a grand scale. Perhaps the real takeaway is that this only hints to a failure of the process than these specific stats. I’d say something like Contact% is still immensely valuable to track but only in light of other information.
I’d say the lesson I learned from this is to look more frequently at OBP and ISO as my main indicators of a players fantasy value. If a player is showing well in both of them and the fantasy numbers aren’t there yet, they will be soon enough.
Erudite Fantasy Baseball 