On Fantasy Baseball Trading
published May 18, 2005
by David Luciani

An overwhelming majority of recent questions submitted to my "Ask David" email inbox have been related to potential fantasy baseball trades.  I'm not just talking about trades in the sense you might think as a trade is simply an exchange of one or more players for another player or players, meaning that every time you drop a player from your roster and pick up a replacement from the free agent pool, you are in essence making a trade.  Whether you're dealing with a human trading partner or have your selection of the players no one else has picked, trading is a key ingredient of fantasy baseball success in virtually every league in which you will play.  This week, I'm going to focus specifically on trading, both with real opponents and with the free agent pool, and I'm going to incorporate examples from the actual questions you all have sent for consideration for the column.

Several times, preparing for this column, I went through the questions sent to me related to trades and I see, unfortunately, that there are still a number of misconceptions out there about how to properly value players, particularly in trades at this crucial time of the year.  Mid-May is one of the most crucial points of the year because players you acquire now will still accumulate the bulk of their performances for your fantasy team if you acquire them and it's early enough that performances to date can be misleading.  Fortunately for the smart fantasy player, it's late enough that many people consider year-to-date performances to be meaningful enough that they will often give up on a player because of how he's done in the first six or seven weeks of the season and so the time for exploitation of opportunity is never better.

One of the most common understood techniques for successful trading involves the "buy low, sell high" mentality that has pervaded the fantasy baseball thinking population.  Regardless, there are people who still don't quite get the idea and more importantly, who exaggerate the concept to be something it isn't.  I recall reading somewhere (I think it was one of the official Rotisserie rule books long before I started being a contributing writer to it) that you should "trade them while they're hot and get them while they're not."  What some questions I receive clearly demonstrate is that some readers make the mistake of presuming that if a player started slow, that he's going to somehow overcompensate with tremendous performance the rest of the way so he finishes with his usual numbers.  Case in point is demonstrated by a reader who sent me this question, which was really more feedback but suits my purposes here of demonstrating a misconception:

"I've been offered Eric Chavez in a deal involving multiple much weaker players on my end.  It just shows that people out there are quick to panic and don't realize that Chavez will get hot and hit .320 the rest of the way and then finish the year with his usual .270 or .280 average."

I appreciate the reader's comment here but in fact, what the reader is demonstrating here is what is called the "gambler's fallacy" and it's the biggest mistake that smart fantasy baseball players can make.  That is, we're talking about those fantasy players smart enough to recognize that you should acquire players who are performing below their real level and trade overachievers.  The gambler's fallacy, which you can read much more about by searching online, basically is the idea that if an event has not performed at its real ability or level to date, that somehow it will overcompensate later by being "due" to perform.  So, a roulette wheel is spun five times and comes up with a black number five times.  The fallacious gambler decides that somehow red is "due" because the odds of six straight black spins is remote, at best.  The mistake made is that while the odds are indeed low of getting six black spins in a row, the odds are low if you're attempting to predict it before any of the spins are made.  When the first spin comes up black, now you're only assessing the odds of five straight black spins and when the second comes up, now you're only looking at the odds of four straight black spins and so on.  When five spins in a row have already come up black, then that sixth spin has the same chance of coming up black as it does red because the roulette wheel has no memory of the previous spins.  I don't think I can explain this as well as many materials and Internet sites have out there so if you want to read more or don't quite grasp the concept, I strongly encourage you to search for the "gambler's fallacy" online as understanding it will help you in fantasy baseball valuation of players.  As it applies to baseball, the same concept causes problems.  The terrible portion of Chavez's season is already in the books.  When considering a trade, it should only concern you to the extent that you believe it represents a real deterioration in ability.  If a player hits .200 in April and seems to have hit .300 every year, it doesn't mean that somehow you should believe that if he really is a .300 hitter that he will hit better than .300 the rest of the way to give him his typical final numbers.  It can and often does happen that he gets hot later and it balances out but it's not proper forecasting and it's a misunderstanding of the so-called law of averages.  Believing that Chavez will perform above his real ability in order to balance out his stats would be an absolute example of the gambler's fallacy.  Rather, if you believe he's a .300 hitter, you should project him to hit .300 the rest of the way and accept that the .200 portion of his season makes him less likely to finish with his usual .300 average.  In the same spirit, if he hits .400 in April and you believe he really is just a .220 hitter, then you should project him to hit .220 the rest of the way and accept that the .400 portion of his season is going to ultimately inflate his final apparent numbers because if he really is/was a .220 hitter who got off to a .400 start, then he was lucky in the portion of the season he hit .400.

So, coming back to the Chavez idea, I do like the idea of trading for Chavez but don't falsely believe that somehow he's going to perform above his ability to make up for the portion of his season when he clearly performed below it.  No doubt, he can have a lucky good stretch just as much as he can have a lucky bad stretch and that can level things off but it's not smart forecasting.  Chavez's slow start should be fully expected to cost him in his final numbers and if you're trading for him, you should only be concerned with the portion of the season you're going to be getting and that's the rest of the way.

This really, then, speaks to the many questions I've recently received about potential trades involving Eric Chavez, Vernon Wells, Greg Maddux and Randy Johnson, though we now need to consider such players in two groups.  There are players like Chavez and Wells who have solid, recent track records and whose projection I have openly stated have a decent or good degree of confidence.  Those players were very strong players in either 2004 or 2003 and they've gotten off to terrible starts.  Then there are a second group of players for whom I have forecasted performance above their past few years of performance level.  No doubt two of the most contentious projections this year were those for Andruw Jones and Greg Maddux and this now gives me a chance to speak to the many readers who have written about both.  Even just a week or two ago, I had readers writing to tell me that they were going to trade Jones or already had because they had concluded that the forecast had overestimated his abilities.  No doubt, if you believe that I've overrated a player, then you should absolutely value him only at the level you see as being correct.  I've often said that the materials you see at this site or any other information you gather should always be an exercise of supplementing your own knowledge and not replacing it.  Ultimately, if you win your fantasy league you want it to be because you used all the information you had available to make an informed decision about a player or players and then you pay the consequences or reap the benefits of your decisions.

So, let's deal with Jones before Maddux because they're the two players who have most been mentioned in reader emails.  With Jones, players who were too quick to react or who didn't believe the projection may have suffered a bit because his perceived value has gone up the past week (he's actually now just about on the pace, if not ahead of it, of our Opening Day forecasts, which is amazing even to me because he suffered such a miserable stretch in April) but just because you didn't believe the forecast, means you should have just given him away.

I often argue that the very minimum you should get in exchange for players should be their track record value, particularly their recent track record from 2004 or 2003 and possibly 2002 in rare cases.  That means that if you doubted (or still doubt) the forecast for Andruw Jones less than three weeks ago, when he was hitting just .182 with 2 home runs going into the final three days of April, it doesn't mean that you should doubt it to the extent that you simply give him away, believing he's a .182 hitter.  I can't tell you how common this theme was in reader emails - I had one reader who actually wrote to tell me that I was "so obviously wrong about Jones that (he had) traded him for Nook Logan, just to get anyone else in that useless spot in (his) lineup."  That's unbelievable.  That's a reader who not only abandoned his plan, because he obviously trusted the forecasts a month earlier when he drafted him, but it's a reader who valued Jones not only below my projected level but well below his track record.  This is a reader who, I can safely say here in this space, who is going to lose this year no matter what materials he has available.  I hate to be so harsh and I know it's going to cost me that reader to say it but it's a reality that I must state to make a point for the benefit of the other readers who have the potential, but haven't yet, gotten too emotional.  Granted, not even the most optimistic of Jones supporters (and my projections have clearly shown that I am one of those) could have predicted just how quickly he would get hot but if you thought I was wrong, then you needed to replace my opinion with your own or with something else.  What was your own projection for Jones when you traded him for Logan?  He's averaged more than 30 home runs the past seven years, has had between 90-120 RBI each of the past five, has not scored fewer than 85 runs in a season since 1997 and is just twenty-eight years old, so he can't be in a decline.  The trade you made is one that many leagues would veto for being too unbalanced against you.  My point here, as it not only applies to Jones but to projections in general, is that if you disagree with a forecast, particularly a contentious one that projects an unusual ability given a player's track record, then by all means substitute your own projected value but go to the trouble of doing that and assigning a projected value!  If you think I'm wrong, don't simply dump the player because you assume he has no value.  Assign some projected value of your own and use that as your base.  Never trade a player for value less than his recent track record and you won't feel so bad if he suddenly reverts to his typical levels of performance.

In that respect, we have Greg Maddux, who unlike Jones is not on track to achieve his projection so far... or is he?  I have readers who keep asking me since his third start why my projection has continued to be so optimistic and this is a good time to come back to what I explained about the gambler's fallacy.  If I'm projecting a player to have a low 3's ERA, as I have been for Maddux, it is only for the remainder of the season.  Since that second projection update of the season, I've been saying that Maddux will have a low 3's ERA the rest of the way, not in total.  It may surprise you but in fact, since those first two starts of the season, Maddux's ERA is 3.34 in the six starts he's had since.  Really.  You see, as I alluded to recently, the earlier a player has a bad or good game in the season, the more it clouds his apparent seasonal performance.  With Maddux's current ERA now at 4.15, readers might perceive that he hasn't been performing up to projected ability and no doubt, compared to the Opening Day forecast he hasn't.  But in fact, the projection for Maddux since just that second set of the season has stood up and Maddux has, at least so far in a very limited sample, been performing up to the expectations I set for him for the remainder of the season after those first two starts.  So, what you have to remind yourself, much as I explained with the gambler's fallacy, is that even if my projection for the remainder of the season is fortunate enough to be right, it doesn't mean that you're going to "feel" like the player is having that kind of season if he got off to an unlucky or particularly lucky start to the season.  I am only projecting his ability for the portion of the season that remains not yet played, at all times.

By the way, all of this isn't to crack open the champagne bottle and say that Maddux is going to have that great year.  On the contrary, both those first two starts and the six starts since all remain a sample that can't tell us for sure.  I'm just emphasizing that many of the questions sent to this space ask the question "How come you haven't downgraded or upgraded Player X yet when he continues to have a good/bad season?" when I can generally respond that if you actually look, you'll frequently find that what the player did for the first week or two of the season has no relation to what he's done since.

So, in the case of trading Andruw Jones, I emphasized that the reader should at least get fair market value for a player projection he didn't agree with but that doesn't help so much with a player like Maddux, for whom I've projected his best season (ERA-wise that is, not win-wise) since 2002.  In such cases, there are opportunities here to profit from Maddux's apparently mediocre start.  First, if you have him on your roster, then you usually have to stick it out until it becomes clear that he's not the player we thought.  But if he's out there on another team, you should have no trouble acquiring Maddux at or around his 2004 value.

That really brings me to a key strategy of acquisition.  Much as I said that the Andruw Jones dumper shouldn't have traded Jones for less than his recent value or track record, you can often pick up a player who is projected to be good and you can pick him up at or below his recent track record value.  That means that let's say that you think I'm right about Maddux having a good year.  That doesn't mean that you should be out there trying to acquire him at his projected value.  On the contrary, you should never pay more for a player than you need to and in the case of players like Maddux, you can acquire him at a price comparable for a player who had an ERA around 4.00, just as Maddux did last year.  This helps you in a couple of ways.  First, if the projection turns out to be wrong, you still didn't overpay and you get fair value from him regardless unless there's a complete collapse.  Second, if he does live up to the projection or even comes close, then you profit from the deal you made.  There are countless players out there right now, both on opposing teams and even on your league's waiver wire whose recent track record makes them cheap or free to acquire and whose projection justifies you adding them to your team.

Several readers have admitted that they don't understand what I mean and have asked me to further explain or clarify this business of the perception and how April clouds the numbers.  Basically, game one of the season will always be included in your seasonal stats.  Game two will be included from that point on and so on.  What it does is confuse even veteran announcers and players and managers because they "feel" like a player is having a good or bad year based on how often they see those earlier numbers included in the player's totals.  That 0 for 6 a player had on Opening Day will be included in his scoreboard total every day of the season.  To demonstrate this, let's look at Andruw Jones one more time but instead of talking about his power, let's look at his batting average because that's always one of the biggest influences on what kind of season it "feels" like a player is having.  In terms of baseball value, the value does not change because of the order except perhaps, arguably, if you believe that performance in September is more valuable than performance in April.  But certainly, no one is arguing that a May win is more of a clutch win than an April win and so the order should not change the value of a player.  If we think of a player's season as a deck of cards that can be shuffled with results but, when dealt out, give you the same totals ultimately, whether a player hits that first home run on day one or day thirty makes no difference as long as he hits it.

So, looking at Jones, here's a chart of how his batting average has looked this season so far:

Any reasonable baseball person, watching this season, would instinctively feel that Jones was having a bad season until around maybe Day 26 or 27, at which time his batting average starts to come up to a respectable level.  It feels like he's only had a valuable portion of the season briefly.  But consider this:  What if Jones had done exactly as he had done only in reverse order?  His seasonal performance would be equally as valuable and nothing would have changed but here's how his year-to-date batting average would have looked on each day if we started at the end and work our way back to the beginning - Contrary to what you might expect, he doesn't start with the .300 like average of these later days above but rather, hits more than .400 for most of the "first half" when we look at things this way:

Isn't that remarkable?  These are the same performances only in a different order.  In the first player's case, you'd have announcers talking about how he really struggled all season but has finally gotten hot.  In the second player's case, even as he slumped the same way with an 0 for 26 streak starting around Day 23, you'd have announcers and readers defending him, saying that even with the slump, he's still having a great season because only briefly on Day 29 would his average even dip below .300.  You'd probably be getting people acknowledging that obviously he wasn't going to hit .400 for the year and had to come back to Earth eventually but no one would have been questioning his value at any time of the season, even when the 0 for 26 streak arrives.  Ultimately, you have exactly the same performance but by looking at it this way, we don't allow the performances that are earlier skew our analysis of what the player's ability really is.

I make this demonstration because one of the other common reader themes is you get readers who talk about what a player has been doing "all season" (I can't exaggerate just how often I get this phrase in emails) but what they really mean is what the player's year-to-date numbers have looked like all season, with April numbers influencing them more than May numbers and so on.

In that same vein, when you are making fantasy trades, you must constantly assess the real abilities of a player and should never allow the earlier performances to more heavily influence your decisions than the later performances.  If anything, more recent performance has always been a better indicator of a player's real ability.  When a player becomes hurt in such a way as to influence his performance, the indicators will be recent.  When evaluating a player's ability, last year is often a better indicator than the previous year and so on.

There are many players for whom the window of opportunity is rapidly closing to acquire.  Hideki Matsui, a player one reader compared to eating "fork with a spoon" just a week or two ago (and we published that question and response) is 9 for his last 27 (.333) and in the span of two weeks has raised his average from a season-low .231 on May 7th up to .250 as of games completed Tuesday and it continues to rise.  The power hasn't shown yet but again, no one should trade him for less than his perceived value which remains rightly at the 20-30 home run level he showed last year.  It doesn't mean that he will hit 20-30 home runs but it means that if you're trading him, you need to get that kind of value in return because someone out there will still believe that he's ability level.  Vernon Wells and Eric Chavez remain great bargains to be had.

When looking for bargains, age does have to be a consideration.  It's possible that Tom Glavine and Al Leiter really are over the hill, even though I don't think they are.  With players like Wells and Chavez, you know that they couldn't have possibly hit an instant decline but with Glavine and Leiter, that isn't as certain.  Even when I'm projecting such players to rebound, those are the sort of considerations you have to make.  In one expert's league in which I'm currently playing, Al Leiter was actually recently dropped to the waiver wire and was quickly snapped up by another team.  You have to weigh that in your own forecasting and decision-making.

In terms of selling high, many of the questions I get recently are in relation to three players - Brian Roberts, Jeremy Bonderman and Kenny Rogers.  All three offer different examples of player valuation.  First, with Roberts, we see a player who has clearly performed above Opening Day expectations already.  He has already topped my pre-season power expectation and I have slowly upgraded his power potential without yet accepting that it represents a real, permanent change in ability.  As I demonstrated with the April to May order examples above, he's going to always appear like he's having a good year because that season-to-date performance will always be included but my expectations remain much more moderate than his performance thus far.  However, in response to those readers who have him, I still must emphasize that you shouldn't trade him away for less than his perceived value.  A reader boasted to me in an email:

"I believe in your projections and don't get too excited about hot starts.  Thanks to your philosophy, I was able to trade Brian Roberts right now for Derek Jeter, a player who is valued higher in my fantasy domination sheets."

Well, this isn't a bad trade but I hate to tell you that you could have done much better.  What the reader did here was correct in that he recognized that the projected value for Roberts the rest of the way wasn't as high as Jeter - in his league anyway - but the problem is that the perceived value of Roberts was much higher.  Even if you don't believe in the year-to-date performances of possible overachievers, and even putting Roberts in that category does not negate his clear improvement as a player that I've reflected in the projections, you still should try to get perceived value in a trade, false as it may be.

Coming to Bonderman, his Opening Day forecast remains one of the more unpopular ones but what puzzles me is just how many readers disagreed with it.  It's not that I can't see how good his stuff can be but his track record to date isn't great and I explained previously about how I analyzed his late-season performance of 2004.  Granted, he's off to a good start but there are readers who believe he's better than I'm projecting.  To those readers, again I say, if you believe a player is better than I've projected, then you should always substitute your own opinion in place of my projection.  It is you trying to win your league and only you can make the informed decisions you are capable of making.  I think I'm right about how Bonderman is going to perform the rest of the way but what the emails do reveal is that there are players who, neither because of a track record or solely because of a hot start, have high perceived value because of their potential.  I don't believe it's Bonderman's start causing readers to think highly of him - it seems those who thought highly of him already did before the season started.  What's good about this is that if you're considering trading Bonderman and are one of those who don't value him highly, you can probably get higher value for him than his track record.  It emphasizes just how important perception is, even if it isn't yet proven over a very long run such as a full season.

That brings me to Kenny Rogers.  There are players like Rogers who get off to a great start that is clearly unsustainable.  Kenny Rogers is neither a prospect nor a player lacking a track record.  His history is clear and yet his season-to-date may falsely make people believe that he is better than he is.  This is harder to find and takes lots of exploring but if you're playing in a league with any decent number of fantasy teams, you will find someone out there who, while not willing to pay top dollar for Rogers, is going to believe he can sustain a mid 3's ERA the rest of the way.  Again, his season-to-date is going to cloud the issue for quite a while but he is absolutely a player who, if you have him on your roster, you must trade him quickly.  It's not that he won't be of some value but you can upgrade that value right now because of his perceived performance and you will never get more for him.  His track record and age so far outweigh what he has done that there's no doubt that this can't be held up over the long run.

Finally, there are players who may have really realized potential but who are still overachieving.  Take Brett Myers and Clint Barmes.  Myers is showing strikeout stuff beyond his previous years and to many, looks like this year's version of Ben Sheets as a player who has finally put it all together.  I actually think it won't last but let's say you disagree and think that this is his year and his time.  That doesn't change that his performance to date cannot be sustained no matter what.  That is, a player can suddenly become a great pitcher but they don't instantly become a pitcher who can sustain a 1.63 ERA over a full season.  Fortunately, people get so excited about finding the next big thing that they sometimes don't think clearly and they think a Dontrelle Willis or Myers is a "must" to acquire to help their team ERA.  Even with players who I think are of good value, like Willis, or players who I think are overachieving quite a bit, like Myers, in either case, the ERA that starts with a "0" or a "1" is so exciting to potential trading partners that you can often upgrade a Myers or Willis into a player who really can sustain a mid 2's ERA.  In one league I'm in (that doesn't have keepers), I just traded Dontrelle Willis for Randy Johnson and Miguel Batista.  I may be wrong and The Big Unit and Rivera may be at or near their ends but I definitely upgraded based on the percentage play, no matter how it works out.

Clint Barmes falls into this category.  While I know he is on many teams because he had high Opening Day projected value, he is not going to hit .380 or .390 for the next few years and that means that even as highly valued as he remains on many fantasy domination sheets, you can actually upgrade his value right now by finding someone out there who thinks that his ability mixed with the thin air of Colorado is going to make him a .380s hitter for the long run.  It isn't.  He's a player who played in the thin air of Colorado Springs last year and even at the less competitive Triple-A level, hit just .323 in 530 at bats.  He's good but he's not that good and so upgrade opportunity presents itself based on, like Myers, an unsustainable newfound apparent ability.

Whether you're considering which players to pick up off the waiver wire or are talking trade with a fantasy partner, remember the points I have made here.  Don't let earlier performance cloud your evaluation of recent performance.  Don't sell for less than a player's perceived value is, even if you think he's worth less than that.  Never pay more for a player than you believe the general perception is of his value, even if you think he's worth more.  Always try to line up deals involving your overachievers.  Constantly seek opportunities to grab players whom other teams are ready to give up because of a slow start.  Never get too high or too low during the slumps because all they are are clusters of performance types that don't give you the overall picture.  Remember that the order of events always falsely influences perception.  Above all, even when you disagree with what I think a player will do or what any of your other sources think they will do, never simply presume that they have no value just because you don't agree with the forecast.  In every league, there's someone out there who has at least remote interest in every one of your players.  Make them pay and use this time of year like your second draft as you fill your roster with bargains.  Don't be satisfied with what you have and you'll constantly be pushing the standings, no matter where you sit right now.