Danny Santana: Projecting Regression and Estimating Value

There has been occasional to frequent mention of the “fluke” seasonal performance of certain players; namely, Danny Santana and Eduardo Escobar- Brian Dozier in the not so distant past. Usually, the notion is raised within the contexts of position placement and possible trade packages. One recent thread in particular discusses, in depth, the idea of trading Santana for starting pitching help of varying degrees of quality.

http://twinsdaily.com/topic/15926-mets-need-a-ss-twins-are-a-viable-trade-partner/

When a young player, or a historically average player, performs better than expected it is certainly prudent to keep our expectations for future performance tempered. But, is it mutually wise to expect significant regression? To what extent is significant regression projectable? We have the tools of metrics and scouting reports available to help in such endeavors, but is their sum enough on which serious team and organizational decisions can be based?

First, within the contexts of starting lineup/player position and possible trade packaging, I think you have to acknowledge the value of the player based on his most recent season’s performance. What is the value of a SS who plays solid defense, and hits .275 with 35 doubles in 133 games? What is the value of a CF who plays solid, improving defense, runs plus, and hits .320 with 70 runs and 41extra base hits in 101 games? Obviously, the value is excellent. In combination with Brian Dozier at 2B, it might be the best up the middle value in the league (even without considering payroll). In the case of Santana, I would say that that offensive line in combination with his ability to play the two most premium defensive positions on the field, make him the most valuable player on the team. I would not trade him for any pitching prospect. In the case of Eduardo Escobar, his ability at SS, in combination with his offensive stat line, I think is good enough to stick at SS to open the season and keep Santana out of his natural position and in CF- although, this not happening would be far more palatable to me than trading Santana for a SP. Without getting into too many other factors and variables, I understand that Santana is probably the team’s SS of the future, and it would be perfectly reasonable for Molitor to want to start him at SS. After all, he can’t be a SS stop unless he plays SS.

But I have veered slightly off course, become slightly carried away at the notion Escobar and Santana will continue to rake so prodigiously. I have wandered from my own warning: prudence begs temperance.

.405

The BABIP of the MLB categorical leader, who is Danny Santana. Damn, that’s high. BABIP is probably the number one metric for projecting BA regression. BA rises and falls with BABIP and a BABIP this high is not realistically sustainable. Over the past three seasons, only four players have recorded a BABIP of .390 or higher, eleven at .380 and up, eighteen at .370 and up. Discounting his 24 games at AAA in 2014, Santana’s highest MiLB BABIP was .353- a high number, but not quite four hundo (there is a steady rise in both BABIP and BA from level to level for Santana). According to this logic, we can expect to see a drop in BA for Santana in 2015.

But while .405 might be exceptionally high, does it mean that Santana will not be able to maintain a more reasonably high BABIP, in the vicinity of, say, .350 plus? Over the past three seasons, on the list of the top 25 players in BABIP (.340-.400) are some regular names; some are exceptional: Mike Trout, Yasiel Puig, Joe Mauer, Andrew McCutchen, Paul Goldschmidt; some are not: Dexter Fowler, Chris Johnson, Starling Marte (with the exception of Chris Johnson, all of these players have a MiLB track record of hitting for high BABIP). So, hitting for a high BABIP can be done with some measure of expected consistency. Additionally, very few of the players on the lists were outliers- without previous seasons of high BABIP (Jhonny Peralta in a shortened 2013 and Jer Saltalamachiado also in 2013 are two examples).

Of course, the success of the aforementioned players bears little influence on that of Danny Santana. But there are additional batted ball statistics in support of Santana’s ability to maintain a high BABIP. The type of batted ball that results in the most hits is the line drive (this is actually an assumption, I haven’t looked it up). Santana’s LD% of 26 is excellent, 14th in baseball. His GB% of 45.9 puts him in the high middle, and his FB% of 28.1 puts him in the low middle. Fly balls that don’t result in home runs usually result in outs. Santana’s percentage of home runs to fly balls was on the low side, 8.5%, so the proportion of ground balls he hits to fly balls I think is appropriate for the type of hitter he is. I didn’t find average fly ball distances for Santana, but I think that would be helpful in projecting his future home run production- whether or not we might see more balls travel over the fence or more balls fall short for outs in 2015 (he’s listed at 5’11 160. 160 is pretty light for 5’11. It seems like he could easily put on 15 to 20 lbs without losing speed, and actually add some distance to those fly balls). The other significant stats here are his infield hit total/percentage and his bunt hit/percentage, and both are pretty good. He had 17 IFH with a percentage of 12.7 (the percent is relative to ground balls) and 9 bunt hits in 18 attempts. These, to me, seem like projectable statistics because of their relativity to a player’s speed (which is of course relative to his health).

Now, if I was a little savvier, or a little more motivated, I would figure out how to put Santana’s spray charts and heat maps from both sides of the plate into this piece right here:

http://www.fangraphs.com/zonegrid.aspx?playerid=10322&position=SS/OF

http://www.fangraphs.com/spraycharts.aspx?playerid=10322&position=SS/OF&type=battedball

The links will have to do. The spray charts illustrate his ability to hit line drives to all fields from both sides of the plate. The power is all pull, but relative to his ability to hit for high average it’s the line drives that count. The heat maps illustrate some susceptibility high in the zone, mostly as a right handed hitter, but overall, little weakness.

And here, check out his highlight video:

I don’t see anything mechanically to pick on. The swing is balanced and strong. There doesn’t seem to be any exploitable timing mechanism. His hands are still, save a slight movement back in the load, and they stay in to the point of contact, finishing high- evidence of that slight Ted Williams uppercut. I love to watch his feet the most. They also are quite still, starting wide in the stance and foregoing a stride but still generating power with a back leg load and forward weight transfer so smooth it’s almost unnoticeable. If you can freeze frame on the point of contact on one of his homers, it’s picture perfect. The homers are definitely not moon shots, but they’re not all wall scrapers either. My favorite of all of these are his steps out of the box. His swing finishes so balanced that he is very efficient coming out of the box. It’s almost identical whether he is hitting a GB single or a dinger- he plants his back foot and his first step is down the line toward first. I think this speaks to his infield hit totals as well as stretching singles into doubles. But he is not sacrificing any authority in the way that Nori Aoki does with that hideous running slap swing. Honestly, I got the notion to watch some film on Santana after I started writing this piece, and now that I have I’m even higher on him than I already was. Dude’s sharp; definitely more polished at the plate than I realized.

Granted, I just watched the good stuff. (Well, I did find one clip of a MiLB strikeout, but it didn’t unfound any of the goodness). But it looks like homey will hit, and he has too. Because if he doesn’t, he won’t be getting on base much, as his plate discipline is not so great. To begin with, his O-swing% (% of pitches swung at outside the strike zone) is 40.3, .1% behind our other friend, Eduardo Escobar, ranking them 11 and 12 in the bigs. Unfortunately, as is the case with other hitters who have a high O-swing%, neither Santana nor Escobar have a particularly high Z-swing% (% of pitches swung at in the strike zone) 63.1 for Santana and 67.4 for Escobar. I don’t know about you guys, but I just don’t have the historical standards of these advanced metrics locked into my brain yet (like BA, OPS, SLG, etc) so I have to look at lists and charts to compare. Anyway, having a high O-swing isn’t necessarily bad if you have a high O-contact to go along with it (i.e. Pablo Sandoval, Victor Martinez) but this is not really the case for Santana- his O-contact% was 65.9 which isn’t terrible either. Both his Z-contact and overall contact% end up fairly average (the pitch f/x numbers are slightly more favorable).

Optimistically, and somewhat reasonably, one would hope that as Santana gains experience, his plate discipline skills/numbers will improve and will translate to a BB% closer to league average and an even higher OBP%, which due to his BA is much higher than league average currently. I was about to suggest that with the success of some of the Twins hitters in the plate discipline department (Mauer, Dozier, and Plouffe) that some of that success might rub off on Santana. But then I considered the likes of Oswaldo Arcia and Kennys Vargas and had to laugh at the pretty dramatic (I’ll say coincidental) contrast between the approaches of our American hitters and our Latino hitters. I’m glad we have both. I wonder if/hope Brunansky is able to help individual hitters craft their own personal approach and style to be the best it can be, or if he has the tendency to force players into a single desired approach. It seems like in the past, the staff has been problematically forceful (not necessarily in temperament) with a philosophical bias toward things like patience, seeing pitches, and taking the ball the other way. I certainly think those are good things, but I also noticed that the two world series teams this season had lineups full of free swingers (although I think the team with the more balanced lineup-in these terms- won).

Aside- did anybody else hear Delmon Young’s quote after hitting that bases clearing double against Detroit? They were asking him about it after the game and he said he was trying to hit a grand slam. They were surprised and said something like, “You weren’t trying to stay within yourself and just get a base hit?” and he laughed and said, “No, man, every time I come in I just try to step up and hit the foul pole.” I think that was my favorite quote from the season. I love Delmon.

http://m.mlb.com/video/v36734763/delmon-young-joins-scott-braun-on-mlb-tonight

Haha, okay, so finally, I would like to come back around to Danny Santana’s projected value, both on the field for the Twins and as a possible trade piece. I don’t really understand the Steamer projections (maybe somebody has a great link to a site that explains it) particularly their estimations of games played, which are always really low. For 2015, it seems like they are not very kind to Daniel. They give him about 35 more games (138) without any uptick in counting stats and that BA comes down about 50 points. Ouch. Bastards. For no legitimate reason whatsoever, let’s just split the dif between last season and the steamer projection. Let’s say he plays 150 games, mostly at SS, with steadily improving D. Let’s say he slashes 290/330/420, scores 90, drives in 60, and steals 35. What is the value of that? Let’s say he’s better the next year. What about then?

In conclusion, I would like to see more research and writing done by some peeps more knowledgeable than me on this notion of the “fluke” season. What are some recent historical examples of players having “fluke” good seasons (Chris Davis?)? I probably have overlooked and/or am totally ignorant of other tools used in projecting regression, and it would be good to be educated. Although I am no authority, I do venture to assert that it is not wise to base serious team and/or organizational decisions on the estimation that a player’s recent success of significant sample size may have been a “fluke”. Further, I think that in the particular case of Danny Santana, while it makes sense to anticipate regression, it would be a terrible mistake to overestimate the amount and trade him on its account. I think that both his short and long term value is greater than 90% of starting pitchers.

-All the stats I pulled are from fangraphs.com.