February 19, 2016

The Quinton

On Reaching

by Jeff Quinton

Printer-friendly

Contact Author

It’s the end of outfield week and starting-pitcher week is three days away. There are a lot of outfielders and a lot of starting pitchers. At times, we will be tempted to reach. While we have discussed reaching in combination with other topics, we have not focused on this phenomenon alone. That said, it certainly deserves its own discussion given its prevalence.

My policy on reaching, as you have most likely already guessed, is to not partake. There is probably a little more nuance here and certainly more to be learned, so let us dive into deeper discussion.

First, let us settle on a definition for reaching. For our purposes, we will define reaching as “passing on a player with more expected value for a player with less expected value.” When it’s put this way, we wonder why we would ever do so, but we know the reasons. Usually we reach because the lesser player has more upside than the alternatives with higher expected value, but we also reach for other reasons such as floor, familiarity, consistency, and likeability among other reasons.

After the first few rounds, our options (players) become more and more flawed (relative to the players taken ahead of them), and when faced with uncertainty, we (humans) employ decision-making shortcuts instead of making a decision based on our, or any other, analysis. For example, there is prospect theory, which explains that we choose riskier options when our options fall below our expectations and choose safer options when multiple options exceed our expectations. For another example (and there are so many), we choose lesser players when we are afraid we will miss on a certain tier at a certain position. The uniting part of all these examples is that reaching is just a term we use when decisions get too complex or too scary (and I am not speaking from a high horse, this happens to me all the time) and we therefore frame the decision through something other than expected value. Or, put differently, reaching is just what we do, when expected value tells us to take someone we do not feel comfortable taking.

So we get it, we are flawed decision-makers and that makes us choose options with less expected value than options with more expected value. We have discussed this, the reasons why this happens, and potential solutions at length on these pages. What we have not discussed is when a reach is really not a reach. For example, if ADP and/or the experts have Byron Buxton rated as the outfielder with the 40^th-highest expected value, but your analysis shows him to have the 25^th-highest expected value, then taking Buxton once the 24 outfielders you have ranked more highly than him are off the board is not reaching. It might look like reaching to others, but—per our definition—it is not because you would be taking the player with the highest expected value available. In fact, choosing another outfielder would be reaching because you would be passing on a player (Buxton) that we valued more. It should be noted that this might be bad strategy or a failure to consider game theory, because if we could conclude that our competitors will only take Buxton as the 34^th outfielder overall, we have the opportunity to take some lower-ranked players that our competition ranked more highly and to then take Buxton later. It should then also be noted that while this strategy sounds appealing, it often fails because the players we are “high on” tend to be the players our leaguemates are also high on, which means attempting to game our differences in rankings could lead to us losing the player we thought we could wait on and, more importantly, the player we valued the most.

We get it, valuing players differently than others and then following those valuations when making decisions is not reaching. Another time when reaching is not actually reaching is when our league undervalues replacement level. In other words, depending on owner tendencies, league norms, and league construct (roster size, disabled list rules, minor league slots, etc.), it may be relatively easy to find value on the waiver wire. This would make players with considerable upside (but with greater risk) more valuable than their steadier peers because should they falter, they are easily replaced. The opposite is obviously therefore true for leagues with very low replacement level.

Lastly, there are articles from time to time that ask the question, “When should we reach?” My answer, considering the word should in the question, is never. Always take the player with the higher expected value. The argument against this that gets floated out on the internet is that sometimes players taken later do better than players taken earlier. First, of course, that is how uncertainty works. Second, this is not predictive at all; just because we know that some top 100 players will come from the last five rounds does not mean we know who those players are this year. In fact, the players we think will be those players do not get taken in the last five rounds—there ADP creeps up and up as the season goes on.

One more “lastly” if you will so oblige. There was an article on “when to reach” that concluded that because the average ADP of players that finished as the 81^st-90^th most valuable players in a given season over the past four years was 242—higher than that of any of the following groupings (100^th-110^th through 141^st-150^th)—we should begin reaching in the ninth round. That is neither the conclusion I would draw nor the recommendation I will make. I could understand that being the conclusion if that “242” was the average end of season ranking for players drafted with the 81^st-90^th picks, but that is not what the data was saying. Even if the data was saying precisely that (which it is not), it seems unlikely that it would be anything other than a blip that would be smoothed out over time.

More importantly, what this data is saying is that players we think are bad or not worth drafting end up being valuable every season. My guess is that many of the data points within that “242” are players that were undrafted. Should we therefore reach for players that will not be drafted? No, that is ridiculous. While far less sexy than recommending a time when we should begin reaching in all drafts regardless of context, is that we should be incredibly diligent and active on the waiver wire once the season starts. Reaching is not going to get us those players that come out of nowhere because we are unable to predict who those players are. What will get us those players, however, is staying a step ahead on playing time battles and pop-up players (think David Peralta, Kevin Pillar, Jeurys Familia, or Marco Estrada from last year), while not being afraid to drop players we thought would be better on draft day.

If your opponents decide on draft day that they are going to “strategically reach,” then please feel free to wear a gigantic internal smile while selecting the players with higher expected value that they have left for you.

Jeff Quinton is an author of Baseball Prospectus. Follow @jjq01
Click here to see Jeff's other articles. You can contact Jeff by clicking here