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 40th-highest expected value, but your analysis shows him to have the 25th-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 34th 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 81st-90th most valuable players in a given season over the past four years was 242—higher than that of any of the following groupings (100th-110th through 141st-150th)—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 81st-90th 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.

Thank you for reading

This is a free article. If you enjoyed it, consider subscribing to Baseball Prospectus. Subscriptions support ongoing public baseball research and analysis in an increasingly proprietary environment.

Subscribe now
You need to be logged in to comment. Login or Subscribe
One of the age old arguments we have in our league is how much of fantasy baseball is luck. While it seems like certain managers are in the hunt every year, I wonder if the difference between 1st and 5th is just health and luck.

Is there any way to actually mathematically figure out how much of fantasy baseball winning is based on luck? Or do we just use luck as an excuse when we aren't winning?
My math: when I don't win, the other guy got lucky and I was unlucky. But when I do win, it's because I'm better....
It's sometimes useful to strategically reach.

If the whole league is overvaluing closers then I find it good to reach on a couple so you don't get shut out and wind up with dregs. Just try to reach less than the rest of the league.

Another scenario is if there will be values at one position then reach a bit at another. If for example outfielders are being overvalued to the extent that it's clear good ones will be available late for pennies on the dollar then it's fine to reallocate your resources and reach a bit for a better middle infielder.

This reaching strategy is probably more important in auctions, where you never want to leave money on the table, than in drafts.
In a draft, you want the best player who will NOT be available on your next pick. Even if you had the 2017 BP Annual to guide you (No Uncertainty as to value), you would still be best off using ADP and knowledge of your league's tendencies to include future round availability as part of your draft decisions.

The 2016 BP Annual would have told you that AJ Polluck was a 2015 first round value. You'd have been silly to draft him that early.
Question: assuming you had access to the 2016 BP Annual, where would you have taken Pollock last year if you suspected at least 1 and possibly more of your league mates also had access to it?
You need to consider the opportunity cost (what was your alternative?). If you pass on a similar first-round value player that is 100% likely to be drafted in the first round, then you leave yourself little room to win value. The first round is often a round where you risk losing value because your player does not play up to the pick, not gain value because he overplays the pick.

The other players will also be considering their opportunity cost for taking Pollock, even if they hold a similar opinion to you. At that point, passing in the second round on Pollock becomes gambling on the ignorance of your competitors. Obscurity is not good security in other contexts, and I do not see why it would be in this one. (
"never reach"
Player A is expected to produce $4 of value in 500 plate appearances
Player B is expected to produce $3 of value in 300 plate appearances

I'm going to reach for player B hoping there is a role change, because clearly his underlying skills are better.
But is that a reach? You are even if you can get a player who generates $1 in 200 plate appearances. That is, unless you are committed to running an empty lineup spot for 200 PA if it does not work out.

As a Jose Reyes owner, I know all about covering 200 PA.
Remember, if you already have two 3B and the next available player with the best expected value happens to be another 3B, you aren't necessarily reaching when you pick another player (with a lower overall expected value) who plays a different position you need for your team. Just make sure that player has the highest expected value among available players at that particular position. Fantasy leagues don't operate as optimal free markets, so there can be no assumption that the extra 3B you picked up (because you mindlessly followed the "pick the highest available expected value" axiom) can be traded for equal or better expected value. Just make sure when you fill positions, you factor in scarcity of available talent at particular positions and your needs at particular positions.
"Highest available expected value" does not incorporate variation in categories. If I've covered OBP, runs and RBI and desperately need a player to contribute Batting average, overall value may not be the best choice.

Of course, you could follow the axiom and hope to trade for BA later. My league however, has a relatively low trade volume, which creates a downside to depending on a trade.