Tout Table: Leading Indicators for Roster Management

There’s about one quarter of the season left to play. Roster decisions are always important, but there’s less margin for error as the season progresses. With that in mind, the Touts were asked:

How relevant is recent performance (past 7, 14 and 21 days) when making lineup decisions for fringier players? Are there any particular metrics more actionable than others?

Nick Pollack (Pitcher List, @PitcherList): For pitchers, it can matter a great deal if there are massive changes in their arsenal. Velocity, pitch usage, increased whiffs, etc. can all indicate a legitimate change has been made and can be incredibly actionable. For example, Reid Detmers’ increased slider usage since coming back from the minor leagues in July.

Derek Carty (RotoGrinders, @DerekCarty): Particularly when talking about hitters, recent performance doesn’t hold any predictive value. The only thing it matters for is playing time. Particularly in deeper leagues where simply getting ABs has immense value, recent role is important, and good performance in that role (even if unsustainable) can lock in additional ABs going forward.

Seth Trachtman (NBC Sports Edge, @sethroto): There are small nuggets that can be drawn from recent performance, but overall, the most value from basic stats is simply plate appearances and pitcher roles (GS, holds, saves). More can be drawn from recent advanced stats like pitch velocity, pitch usage, etc.

CJ Kaltenbach (Fantasy Guru, @TheSeigeDFS): Recent performance matters with the younger players more as their lineup positions are more likely to get drastic improvement during the final six weeks of the season so I’m looking for guys who are hot and get them in the lineup before they get the lineup boost that puts them on everyone elses radar.

Ryan Bloomfield (BaseballHQ, @RyanBHQ): Echoing Nick’s comment on placing a heavy emphasis on recent pitcher changes (pitch mix change, movement change, and velocity increases) and acting quickly on those. Reid Detmers is a perfect example. For hitters, the only “recent” metric I look at is plate appearances. Need to always have your thumb on which players are on the field in addition to platoon splits and batting order. I put almost zero stock in recent hitter performance–so much variance in that throughout a season. I prefer to look at rest-of-season projections for hitters instead.

Larry Schechter (Winning Fantasy Baseball, @LarrySchechter): Trying to fill a couple of holes for my AL-only league, recent ABs are critical to find someone who will play some (or play a lot) and who looks like they have a decent track record for possible success. Welcome to my team Mike Masey and Terrin Vavra. Similarly I have benched Jake Meyers as his games played have dwindled, and he just hasn’t hit as I had expected.

Lenny Melnick (LennyMelnickFantasySports, @LennyMelnick): Basically ,any player being considered for my Roster is considered if they are slumping badly (I) 0 for 20 as an example ,yet still in the teams lineup Only a matter of time until the player hits In addition I try to get the backup to the player in a deep lg

Todd Zola (Mastersball, @toddzola): I’m in lockstep with Derek C, with a bit of a subjective addendum. Matchups and history are easily the most relevant factors, but if a player is hitting the ball hard (harder than normal) even in a small sample, I gravitate towards the player. I know the success could be due to facing weak pitching, but hitting the ball hard is still the best way to success (contrary to some poor research I presented last summer). I won’t rely strictly on how hard a hitter is making recent contact, but it’s at minimum a tiebreaker, and sometimes used to take a flyer if I need lightning in a bottle. WIth pitching, I have (as have others) shown recent outings are telling, so I am more apt to go with a pitcher on a roll than a hitter on a perceived streak.

Ariel Cohen (CBS Sports, @ATCNY): I typically look at the past 28 days in general for most stats/trends. For pitchers, I tend to look at the past 3 starts worth of performaces. For hitters – to find some replacement players, I look at ABs over the past 14 days. Playing time is KING.

Fred Zinkie (Yahoo! Fantasy Sports, @FredZinkieMLB): I’ll look at surface stats (AVG, HR etc.) for the past 7-14 days. Not to evaluate skills, but because players who have been playing well are more likely to get their name on upcoming lineup cards. For pitchers, I’m looking at skills and not recent results.

Brad Johnson (Patreon/BaseballATeam, @BaseballATeam): Regarding Derek’s comment, one of the things I focus on all season is using actual scouting to separate out a few “truly hot” hitters from the noise. He’s absolutely right that at the population level, any kind of hot streak is meaningless noise. However, anyone who has played baseball also intuitively knows that there is such as thing as being “locked in.” Our statistical tools still do a poor job of recognizing it, but I can promise you nearly MLBer would tell you sometimes they’re hot/cold for real reasons and sometimes it’s just luck. The Moneyball reflex might be to dismiss the concept of “real hot streaks” because the statistical evidence is lacking. If the last 15 years of statistical development (and especially Statcast) have taught me anything, it’s that if baseball lifers say something is real and we can’t find it in data, it means our data probably isn’t good enough.

Brad Johnson (Patreon/BaseballATeam, @BaseballATeam): In terms of actionable fantasy advice, the above just means to be skeptical of hot streaks. Assuming you aren’t personally scouting, you can still go after hot hitters on the chance they’re “true hot.” The projections are usually going to be more instructive than recent performance, but sometimes desperation calls for desperate maneuvers. As always, matchup quality is the #1 thing to follow when trying to squeeze real value out of marginal players.

Doug Dennis (BaseballHQ, @dougdennis41): I am only looking at usage and hoping to catch lightning in a bottle (typically) because the data sets we are talking about are so small (backwards and forwards).

Scott Pianowski (Yahoo! Fantasy Sports, @Scott_Pianowski): I’m usually focused on playing time, although sometimes a round of stolen bases can be a little bit sticky because it might encourage the player to keep running. As for a batter’s production in a small sample, it’s mostly important in that it could protect playing time. As most of the room has said, it’s got little to no predictive value, but a hot stretch might at least insure someone’s immediate role with a team. With pitchers, I’m generally focusing on walks and strikeouts, as always, and putting an eye on the upcoming schedule.

Chris Clegg (Fantrax, @RotoClegg): When looking at hitters recent performance I think looking at basic stats can be useful. Playing time matters most, but a player is more likely to keep or gain more playing time if they are providing value to their teams in real life. OBP, runs, and RBI can help show some value they are providing to their team. But at the end of the day, playing time trends are what matter most. For pitchers I do think recent performance matters because pitching has a lot to do with mental state on the mound and recent strong performance can keep a pitcher in a good place mentally. As Nick said, arsenal changes are also a factor when looking at pitchers recent performances.

Mike Podhorzer (Fangraphs, @MikePodhorzer): Recent performance over such a tiny sample is never relevant, regardless of the fringiness of the player. The only thing I might care about is a pitcher’s velocity — if it’s up recently and has led to a higher SwStk% and strikeout rate, then I would pay attention.

Tim Heaney (TheGameDay.com, @TeamHeaney): These splits are factors but will hardly tell the whole story. As the edict has been spewed out by many a fantasy tout, past stats only help you so much, if at all. We care about what’s to come. For batters, handedness and stadium in which they’ll hit play the most for me on the fringes. As for pitchers, facing lineups with high strikeout rates typically dictate streamer opportunities, depending on how deep I need to fish for useful arms.

Derek VanRiper (The Athletic, @DerekVanRiper): The volume of playing time, and even batting order locations matter to me in smaller samples, but the actual output is much less important due to the limited value it provides when slicing things into small segments. Within a smaller sample, I’m somewhat interested in seeing process-related numbers for hitters (i.e. hard-hit rate), but I try to limit how much I let performance during a small sample guide decision-making.

Sara Sanchez (bleedcubbieblue.com, @BCB_Sara): I agree with everyone above who said playing time matters more and I almost never make decisions on 7 or 14 day windows. However I do like to look at 30 day windows on occassion to isolate trends for pitchers and hitters. If I notice large changes I will dig deeper to see if pitch mixes have changed or maybe an individual hitter is pulling the ball more. I’m basically looking for anything that looks like a substantive change in a pitcher/hitters approach.

Doug Anderson (Fantrax, @rotodaddy): Projected playing time take precedence over recent results, but I also look at the upcoming schedule and see what ballparks players will be playing in and the pitching staffs they’ll be facing. I think that’s a better predictor of short term success than a player’s results over the last few weeks.

Charlie Wiegert (CDM Sports, @GFFantasySports): Very relevant. Players are streaky! Looking at season long stats, I wouldn’t play some guys like a Max Muncy. As Ron Shandler would say, once a player owns a stat, he’s got a good chance to repeat it. If players have a good history, and haven’t produced to that level, watch for them to get hot…..except for Bellinger!

Jeff Erickson (Rotowire, @Jeff_Erickson): I want to pretend that I’m savvy enough as a player to avoid recency bias, but that would be a lie. It informs playing time, and that’s what matters when you’re looking at fringier players on the waiver wire. That’s especially true the deeper the league is – after all, these players that are available in AL Tout Wars are available because they are flawed players in the first place, so why not err in favor of playing time?

Peter Kreutzer (Ask Rotoman, Fantasy Baseball Guide, @kroyte): I filter the free agent list each week by the last 14 or 21 days of stats. Not because these lists give me an idea of player talents, but because I’m always looking for at bats and innings pitched. There are pitfalls here. When a regular returns at bats for his sub dry up, but in deep leagues finding at bats often means finding value and there usually aren’t many choices. Identifying the best choices doesn’t always lead to success, an understatement, but keeps you in the game when someone does break out. And every once in a while someone does. Late in the season that’s what you’re looking for unless a top prospect arrives.

Jeff Zimmerman (Fangraphs, The Process, @jeffwzimmerman): For hitters, plate appearances, that’s it. For pitchers, 21 days includes 3-4 starts and several pitching metrics can restalize after 3 to 4 starts. I use it to compare to the full-season stats to see if there is a possible change in pitch mix or velocity.

Mike Sheets (ESPN, @MikeASheets): I will echo the group and say that I’m primarily looking at playing time right now. Even if I believe in the skills of a player, it doesn’t matter if he’s playing only 3-4 times per week. That said, recent results do affect playing time on many teams, so it’s worth paying attention to whether a recent hot streak is leading to a more regular role.

Ryan Hallam (Fighting Chance Fantasy, @FightingChance): The later it gets into the season the more the most recent data matters. It is getting close to the end and for instance the H2H league I am in there are only three weeks left for the regular season. A hot streak can last a few weeks and help propel you into the playoffs whether it is a hitter or a pitcher. Conversely, if a guy is starting to tail off it could certainly be fatigue or the fact that their team might not be playing relevant baseball right now. Early in the year I don’t put much stock into the first few weeks, but at the end it matters very much

Patrick Davitt (BaseballHQ, @patrickdavitt): I’m with the group: PAs and batting-order changes for hitters, SwStr and K%-bb% for pitchers, as a pointer to take a closer look at pitch-mixes and pitch-effectiveness changes.

Rudy Gamble (Razzball, @RudyGamble): I concur with the group. For hitters, it is typically looking at recent playing time, batting order, and projected matchups (pitcher quality, ballparks). For starting pitchers, a recent hot streak supported by solid K/BB is a good sign of health and possible improvements. For relief pitchers, recent performance a good gauge for both health and role/usage.

Bret Sayre (Baseball Prospectus, @BretSayreBP): It’s rarely relevant after the trading deadline because the quality of competition being faced is all over the place. For lineup decisions in August and September, I like to look forward to matchups using a season-long, high-quality metric like DRC+/DRA as the baseline rather than look backwards at recent performance.

Vlad Sedler (FTNfantasy, @rotogut): Recent output is part of the equation but it’s always best to look under the hood at advanced metrics, lineup slot/team context (for hitters) as well.