Twisting Oliver: Off-the-radar batters our projections love

Cody Ross will be a top-50 offensive player. Rajai Davis will steal more bases than all but two players. Dan Uggla will be one of 15 players to drive in 100 runs. Nelson Cruz will be worthy of a top-40 pick. Alex Rios will re-emerge from his season-long slump.

These are just some of the bold predictions THT Forecasts (aka Oliver) makes about offensive performances in the upcoming season. Last week, I shared some techniques for getting at this kind of information. Now, I’ll be going a little more in-depth into the kind of numbers we can expect from some players who will likely be undervalued at your fantasy draft. In this column, I’m using a basic formula that gives players one point for every run, RBI, walk and single; two points for doubles and steals; three points for triples; and four points for homers. I also subtracted one point for a caught stealing and another half-point for every strikeout.

One thing to keep in mind during all of this: Don’t get caught up in the actual numbers. Oliver tends to be a little more pessimistic (realistic?) than many of the other projections out there (ESPN has 34 players batting at least .300, while Oliver has just 13). It’s a fool’s errand to compare one player’s predicted level of performance in Oliver directly against one from another source. What matters is relative performances, how a player fares in one projection system against others in the same system. (If you really want to get wild, try taking averages from multiple systems and comparing the averages.)

I know it’s been said many times before, but the key to a successful draft isn’t just the ability to find diamonds in the rough or hitting home runs with every early pick. The goal is to avoid blowing early picks and making the most of the late ones. With that in mind, I figured it was worth going through several levels of players from guys being taken in the first five rounds all the way down to guys who are sometimes going undrafted.

Top of the draft bargains

Cruz: Statistics-minded owners have had their eye on the Rangers outfielder for a while now, but that doesn’t necessarily mean all his value has evaporated. In Yahoo leagues, Cruz is still going around No. 67. ESPN has him ranked as the 19th-best outfielder. Oliver wouldn’t bat an eye if you took him in the top 40, though. With a projected line of 35 homers, 100 RBIs, 13 stolen bases and 86 runs, there are few players who will produce better across the board. Oliver is a little less excited about his batting average (.263), but that’s a small price to pay.

Curtis Granderson: I know what you’re thinking: “How’s a guy projected to go in the 50s supposed to be undervalued?” He produces like a top-20 player, that’s how. Yahoo drafters have been taking him at No. 56 on average. If Oliver’s right, taking him there would be an absolute steal. His 25 homers and 90 RBIs are just the tip of the iceberg. Factoring in his 15 steals and 95 runs, that .264 batting average will be only a slight nuisance.

Ben Zobrist: Obviously, the Rays utility infielder is no longer a super-sleeper the way he was last year, when he went undrafted in many leagues. Still, owners seem to be treating his breakout season with a little trepidation, taking him in the early 40s in Yahoo leagues despite being position eligible at 2B, SS and OF, and ESPN ranking him as the seventh-best 2B option. Amazingly, Oliver deems him a good value even at those lofty heights. My ranking system pegs him as the 13th-best offensive player, two spots behind Chase Utley and seven spots ahead of Ian Kinsler (the first- and third-best 2B options). Oliver likes him for 25 homers, 89 RBIs, 13 steals and 93 runs scored, despite predicting a more than 30-point drop in batting average.

Safe bets in the middle rounds

Uggla: Yahoo owners have been taking him in the mid-70s, while ESPN ranks him as the 12th-best 2B. Generally seen as more of a fallback than leading man, Oliver thinks he’ll be a premier source of power at his position. Oliver says he’ll be the fifth-best 2B overall and could break the top 30 offensive players, while being one of only 16 players to hit 30 homers and one of 15 to drive in 100 runs (and the only 2B to do either of those).

Adam Dunn: His measly batting average constantly drives down his draft position, but Oliver definitely thinks his performance in other areas makes him well worth the hit. ESPN rates him as the 14th-best 1B and he’s being drafted in the mid-70s at Yahoo. Oliver puts him among the top 25 hitters, well ahead of such first basemen as Kevin Youkilis and Joey Votto. Oliver seems to think he’s gotten his strikeouts under control (“just” 159), while maintaining his ability to draw walks (a league-leading 106) and not ruin your batting average (.252). He’s also pegged at 36 homers and 100 RBIs.

Lance Berkman: Just one year ago, the Astros first baseman was being drafted around the same time as Ryan Howard and Prince Fielder. Now, he’s falling into the mid-90s in Yahoo drafts and ESPN ranks him below the likes of Billy Butler and Kendry Morales. Oliver thinks he’ll produce more like the top-10 first baseman to which we’ve grown accustomed. His projected 24 homers and 87 RBIs don’t necessarily jump off the page, but if your league looks at walks (92) or OBP (.387), his numbers start to look a little better. Basically, he’s Pablo Sandoval 60 picks later.

Finally, some legitimate sleepers

Ross: Will the Marlins outfielder be the best offensive center fielder in all of baseball? I have a hard time buying that, but who am I to argue with Oliver? In any case, you won’t have to take him nearly that high in order to get plenty of value. He’s not even being drafted in the top 200 at Yahoo and ESPN ranks him as the 53rd-best OF. Oliver, meanwhile, projects a line of 29 homers, 97 RBIs and 40 doubles, which, using my points system, makes him the 19th-highest-scoring batter. Even using a standard 5×5 measure I created, {R+RBI+(HRx2)+(SBx2)}xAVG, he’s still the 37th-best hitter. Taking a late-round flyer on Ross seems like a no-brainer.

Franklin Gutierrez: The Mariners outfielder made a name for himself by registering one of the greatest defensive seasons in history. Unless you play in a particularly odd fantasy league, though, this is of no value. So, it’s understandable that he’s out of the top 200 picks in Yahoo leagues and ranked as the 47th-best OF by ESPN. Oliver likes him as the 78th-best hitter using my points system and at 72 with that 5×5 calculation. His production in any one area won’t win categories, but a projected line of 20 HRs, 78 RBIs, 14 SBs and 79 runs is pretty much on par with the projections of Bobby Abreu and Raul Ibanez.

Rajai Davis: Owners who were paying attention last year unquestionably remember his second-half performance in which he stole 30 bases, hit .325 and scored 46 runs. Apparently, not many of those people have much faith that it was a sign that he had finally figured it all out as he’s being drafted toward the very end of most Yahoo drafts and is the 48th-ranked OF at ESPN. Oliver scoffs at those numbers and projects robust enough stats to place him at No. 81 among batters in my points system and all the way up to No. 49 in my 5×5 ranking. Oliver suggests 46 steals (behind only Jacoby Ellsbury and Michael Bourn), a .277 batting average and 80 runs, or across-the-board production on par with B.J. Upton.

Rios: If there’s one player who was a bigger disappointment than the White Sox outfielder last season, I can’t think of him. Once considered a possible keeper in many leagues, he’s now being drafted in the 160s at Yahoo and is the 35th-ranked OF at ESPN. Granted, those aren’t bargain-basement numbers. Still, if Oliver’s right, he’s a good gamble. In both my ranking systems, he’s among the top 60 batters with a line of 15 homers, 75 RBIs, 21 steals and 82 runs or the rough equivalent of Shane Victorino about 70 picks later.

Other stray observations

  • One position that Oliver seems rather down on is catcher. Other than Joe Mauer, no one ranks in the top 50 and only Victor Martinez and Brian McCann crack the top 100 batters. If there’s a draft strategy there, I’d suggest grabbing Mauer early or waiting until much later to grab someone like Mike Napoli around the 14th round or Ryan Doumit around the 18th. Both players project at levels similar to Matt Weiters, who’s being picked at No. 38 in Yahoo, and both project much better than Russell Martin, who’s generally the fifth catcher off the board. Believe it or not, Jorge Posada projects as the fourth-best catcher (although I’d say he’s only worth the pick if he falls well beyond his average Yahoo pick of 141).
  • Generally speaking, I’d say Oliver is not particularly fond of drafting speed merchants in the higher rounds. Some of the players our projection system seems to be a bit wary of in relation to where they’re being picked include Ichiro (average Yahoo pick: 29; my points ranking: 58); Denard Span (AYP: 80; MR: 114); and Brian Roberts (AYP: 46; MR: 132).
  • In case you’re curious, Hanley Ramirez is the top player using my 5×5 calculations and Albert Pujols comes out on top when I’m using points.

A Hardball Times Update
Goodbye for now.

22 Comments
Oldest
Newest Most Voted
Inline Feedbacks
View all comments
Ralph Kramden
14 years ago

In case you’re curious, Hanley Ramirez is the top player using my 5×5 calculations and Albert Pujols comes out on top when I’m using points.

Do your calculations utilize VORP?  Are these guys as valuable compared to other guys at their positions as Mauer is at catcher (even though Mauer might not score as high in absolute terms)?

Jonathan Sher
14 years ago

Who are you top 10 for 5×5?

Jeremiah
14 years ago

I made no attempt to compensate for position, which is probably a good project to undertake for future columns. Although, I will say I think it’s easy to give too much attention to position scarcity, especially early in the draft. I also didn’t attempt to compare the value of hitters to pitchers. Definitely still trying to refine my ranking system.

My top 10 for 5×5:
1. Ramirez
2. Pujols
3. Braun
4. ARod
5. Miguel Cabrera
6. Adrian Gonzalez
7. Prince Fielder
8. Ryan Howard
9. Mark Teixeira (yeah, there’s a few 1Bs…)
10. Matt Holliday

Ralph Kramden
14 years ago

I *always* apply VORP in my roto drafts.  (I also almost always win my roto leagues.) grin

Mike Podhorzer
14 years ago

I don’t get it. You wrote an entire article looking at who Oliver loves and then applied some random point system that no fantasy league uses and don’t even adjust for position.

And you keep mentioning how Oliver is ranking a player. Oliver is just the projection system, your valuation method is ranking the player!

So I really don’t understand the point of this article. You’re trying to give fantasy advice based on the rankings from a valuation method that is clearly flawed. It’s no different from me creating my own valuation method that awards 10 points per batter strikeout and then claiming Mark Reynolds should be the 1st overall pick and is therefore undervalued!

Jeremiah
14 years ago

I guess I just disagree. My valuation system(s) are pretty straight forward. One utilizes an extremely straight -forward points system not unlike many leagues in which I’ve played. The other is a slight twist on 5×5. Either way, you need to figure out a way to place a value on Oliver’s projections. You’re right, Oliver doesn’t “rank” players and maybe I could’ve made that more clear. But my rankings are entirely based on Oliver projections of commonly used fantasy stats. If you have a better way of projecting players, I’d love to hear it. I’m in no way suggesting mine is the only, or even best, way to do it.

Mike Podhorzer
14 years ago

Well let’s see, you arbitrarily choose to multiply both home runs and steals by 2. Then you decide to multiply everything by batting average? That makes absolutely no sense. There isn’t even a factor for at-bats. So you are saying a .300 batting average carries the same value if posted over 200 at-bats as 600 at-bats, which is clearly wrong.

You also cannot compare your rankings derived from your random points system to the ADPs on Yahoo. The Yahoo leagues the ADP list is based on does not use your points system. It uses a 5×5 roto format. Rankings, and as a result ADPs, will change, and sometimes dramatically, depending on the scoring format. So you’re comparing apples and oranges here.

Jeremiah
14 years ago

The HR and SB multiplier is random, no doubt, but it at least attempts to regulate for the scarcity of those stats in comparison to RBIs and runs. I definitely could have chosen another number, but it would have still been relatively arbitrary and ran the risk of overcorrecting for their relative scarcities. Any rate stat is going to have to be factored into a ranking system somehow and since avg is the most commonly used one, that’s why I chose it. The idea that a 200 at-bat player gets equal treatment as a 600 at-bat one, though, just doesn’t hold up. No player with fewer than 500 projected plate appearances appears in my ranking’s top 75 and after that the number of plate appearances remains mostly inconsequential.

As for using Yahoo’s ADP, it’s merely as a way of showing where players are being drafted. The reality is players go in different places in all leagues, I can’t possibly regulate for all of them and I don’t try.

All I can do is be transparent with my ranking system, which I believe I’ve done. I’m sorry you don’t like my chosen methodology (although I’m sure you disagree with my use of that term, as well). In any case, I invite owners to create their own systems that are specifically catered to their leagues and even outline ways for people to do that. I tried to use a pretty generic system (or at least utilize stats that are commonly used).

What this article attempts to do is to show how players stack up within Oliver compared to others in the same projection system. Hopefully, that was clear enough.

Jeremiah
14 years ago

I considered the same thing, but since Oliver projects just 22 SBs, 190 hits and 80 runs for Ichiro (all career lows); 20 SBS, a .284 AVG and 77 runs for Span (all below the projection systems you mention) and 20 SBs, a .276 AVG and 67 runs for Roberts (again, way below other systems), I felt like this was not going out on a limb for them.

Using Fangraphs Fans projections, Ichiro ranks 39 in points, Roberts is 36th and Span is 58th (all much higher than in Oliver). Using my 5×5 standard, they rank Span is 61st; Ichiro is 18th; and Roberts is 40th (again all way higher than in Oliver).

I definitely could have mentioned that, but I do think it shows Oliver disliking those players more than it is my ranking system.

BobbyRoberto
14 years ago

Using the Fans projections, which are notoriously optimistic, is misleading.  These are the projections for Ichiro (I ordered them by projected PA):

.335, 710 PA, 114 R, 9 HR, 71 RBI, 32 SB—Fans
.319, 701 PA,  95 R, 9 HR, 50 RBI, 27 SB—Bill James
.313, 648 PA,  85 R, 8 HR, 46 RBI, 23 SB—CHONE
.328, 623 PA,  80 R, 7 HR, 62 RBI, 22 SB—Oliver
.319, 614 PA,  81 R, 8 HR, 45 RBI, 27 SB—Marcel
.325, 610 PA,  82 R, 7 HR, 44 RBI, 24 SB—ZiPS

You compared Oliver to the Fans, which shows a huge difference.  Compared to CHONE, Marcel, and ZiPS, there is very little (if any) difference.  In fact, Oliver’s projection for 62 RBI for Ichiro looks to be quite optimistic.

Jeremiah
14 years ago

This tells me that generally speaking Ichiro is being overdrafted. Most of those numbers don’t paint a particularly bright picture of Ichiro’s performance. I only have the spreadsheets for Fans projections handy (why I compared him there). Just looking at those numbers, I’m guessing Ichiro would rank pretty low no matter what ranking system you use.

Again, I’ll put this invitation out there: If you have a specific ranking system you’d like to see used, I’d definitely like to see it. I never claimed that my ranking system is perfect, but the fact appears to remain that Oliver is not particularly bullish on Ichiro’s ability to do anything but hit for average.

Jeremiah
14 years ago

So, I’m running my ranking systems in Chone, Zips and Marcel as well as running the 5×5 for those three players, here’s what it’s spitting out:

Ichiro     Pts   5×5
Zips       72     63
Marcel   50     38
Chone   55     52
Oliver     58     35

Roberts   Pts   5×5
Zips       19     25
Marcel   16     25
Chone   21     32
Oliver   142   130

Span     Pts   5×5
Zips     72     68
Marcel   49     34
Chone   82     77
Oliver   113   102

Looks like my most egregious mistake wasn’t so much the calculations I chose to use, but including Ichiro in this group. Oliver is much harsher on Roberts and Span than Ichiro and if I’d done the 5×5 calculation for Ichiro, I would have saved myself some worry.

That said, I thank Bobby for giving me a good way of proving the validity of my calculations.

BobbyRoberto
14 years ago

Jeremiah, you wrote:
“What this article attempts to do is to show how players stack up within Oliver compared to others in the same projection system. Hopefully, that was clear enough.”

My question is this:  did you use your ranking system with some other projections, such as Chone or ZiPS, to see how they rank players (again, using your ranking system)?  My take is that you created a system to rank players and used it with Oliver projections, then compared the rankings to Yahoo or ESPN’s rankings.  This doesn’t tell us if it’s the projections or your system that causes the difference in rankings.

You also wrote:
“Generally speaking, I’d say Oliver is not particularly fond of drafting speed merchants in the higher rounds.”

To show this, you need to use your ranking system with some other projections to see where the “speed merchants” rank.  If Oliver, Chone, ZiPS, Marcels, etc. all rank “speed merchants” lower than Yahoo’s ADP based on your system, then it’s not likely the projections ranking them low, it’s your system.

I don’t mean to be critical, just trying to point out where this article could be improved.

David Kulich
14 years ago

I’m sure you guys know this, but just throwing in my 2 cents.  Using Ichiro was a bad choice simply because projection systems annually can’t project him.  Look at him for the past five years, and the same drop-off in numbers has been projected for each and every one.  He consistently beats his projections, and that’s why his ADP is higher than what the projection systems say.

Jeremiah
14 years ago

Yes, I’m definitely regretting using him for multiple reasons. I suppose Roberts is another guy who beats projections regularly. Of course, my whole point was that Oliver wasn’t particularly fond of either player not necessarily that those players will be bad this year.

Jeremiah
14 years ago

Just because this is the kind of guy I am, I created a couple more ways to rank players that you may or may not find interesting and I’ll probably end up writing about later either way. Both strive to rank players in standard 5×5 roto leagues and both rank players from 1-417 (the number of batters Oliver projects) in each category.

The one I liked drops whatever ranking is lowest and then gives you an average ranking based on that. The top 5 are Pujols, Braun, Ramirez, Miguel Cabrera and Adrian Gonzalez. Cody Ross is 53rd; Nelson Cruz is 13th; Granderson is 18th; Zobrist 22nd; Berkman is 41st; Dunn is 56th; Rios is 59th; Gutierrez is 61st; Rajai Davis is 88th (two spots behind Span); and Roberts is 135th.

I dropped the lowest ranking because if you don’t the rankings get quite a bit more counter-intuitive: Ramirez jumps to the top (not counter-intuitive I realize); Pujols drops to 27th; Shin-Soo Choo ranks 11th from 30th; Adrian Gonzalez drops to 34th; and Ichiro drops to 60th. Believe it or not, though, Ross jumps to 40th.

In any case, I was actually surprised at how well my original 5×5 ranking held up, at least in the sense that it illustrated the point I was trying to make.

Hayves
14 years ago

“he idea that a 200 at-bat player gets equal treatment as a 600 at-bat one, though, just doesn’t hold up. No player with fewer than 500 projected plate appearances appears in my ranking’s top 75 and after that the number of plate appearances remains mostly inconsequential.”

If you want to check this, compare Polanco’s numbers to someone with similar projections (including HRs, SBs etc) but a fraction of the ABs…maybe someone like Matt Diaz?

“Oliver tends to be a little more pessimistic (realistic?) than many of the other projections out there (ESPN has 34 players batting at least .300, while Oliver has just 13).”

Also, this statement here seems to be the reason that most of your sleepers were chosen as such. Seems to me that BA is going to be even rarer than HRs/SBs this year, meaning that most of your sleepers, who have low-ish BAs will have their value driven down by people like Ichiro who will carry their teams. It would also explain Ichiro’s “overdrafting”, which may not be the case.

Derek Ambrosino
13 years ago

Ralph,

I don’t think the concept of VORP is that simple when it comes to fantasy.

First, how are you defining “replacement” level? What you can get on the waiver wire? What you can get 50 picks later? 100 picks later? Simply the value between the best and next best player at that position? I don’t really understand what VORP even means in this context.

Second, in MLB there will be 30 starting catchers and 30 starting 1Bs on teams’ rosters. This is not the case in fantasy baseball. This affects the baseline value/production of each position. For example, few teams are going to draft more than one catcher (in a single-catcher league), but it’s entirely reasonable that a team may start 3 1Bs (1B, CI, util). So, in your generic 12-team mixed draft league, the worst thing that can happen is that you wind up with the 12th best C, whereas it’s possible that every starting 1B in MLB is owned.

So, even if we adopt the most liberal definition of “replacement player” possible – what is left on the waiver wire. (I’d argue that this is a rather foolish way to define “replacement player,” by the way), for a catcher that might be the 13th best catcher in (fantasy) baseball, while for a 1B that might mean the 31st best.

…Now, you might actually argue that your 1B replacement would most likely come from your own team and not the waiver wire (as you are likely to own more than 1). Then, the replacement player level would be upgraded to the best unowned player on the wire, if you are able to jigger your roster to accommodate him.

This wrinkle makes things a little more complicated, but that’s also s hint that VORP is a somewhat problematic approach to analyzing these things.

In actuality, you are dealing with a constantly changing system, in which there are certain positions more strongly represented at specific points in the evolution of that system. To the extent that VORP exists, it is fluid pick to pick anyway.

This is something many people ignore when talking about position depth.

Jeremiah
13 years ago

Which is just another reason I almost always draft “best player available” unless I need a player just to fill a position. Obviously, having a player like Utley who can play 2B is great, and having someone like Mauer is worth even more, but at the end of day, those guys are putting up great numbers regardless of position. You allow yourself to get too caught up in in drafting guys just because they play a relatively shallow position, you start getting in trouble.

Rudy Gamble
13 years ago

Jeremiah –
While position scarcity is overrated to a degree (except for Catcher), it is definitely a factor that has to be considered.  It’s the reason why you have so many 1Bs in your top 10 (it’s the deepest position) and no Utley.

I also agree with Mike P that using a points-based method is flawed.  At least use the one Tango derived for his projections contest:
Hitting: HR + SB + (H – .27*AB) + R/3 + RBI/3
Pitching: 2*W + SV + K/5 + IP – (H + BB + ER)/2

Derek –
You’re definitely right that VORP is tougher for fantasy baseball.  I moved away from VORP to using an average for the stat in one’s position for several reasons explained here:
http://razzball.com/2010-razzball-projected-point-shares-ver-1-12-team-mlb/

Rudy

Jeremiah Oshan
13 years ago

Fair enough suggestions. I’ve already moved away from the ranking system I derived, but don’t see any harm in using the one your propose for future columns. In my defense, Utley comes in at No. 13 even using my original ranking system, no more 1b’s appear in that top 25 and Mauer clocks in at No. 21. I didn’t mean to suggest that if I was drafting a team, I would do so based strictly on a literal reading of these rankings. Obviously, there would have to be some manual adjustments. That said, I understand the criticisms and have taken them into account.

Jeremiah Oshan
13 years ago

Using that formula for hitters, this is the top 10:

Ramirez, Hanley
Braun, Ryan
Pujols, Albert
Rodriguez, Alex
Cabrera, Miguel
Gonzalez, Adrian
Fielder, Prince
Kemp, Matt
Howard, Ryan
Holliday, Matt

Not that different than mine:
1. Ramirez
2. Pujols
3. Braun
4. ARod
5. Miguel Cabrera
6. Adrian Gonzalez
7. Prince Fielder
8. Ryan Howard
9. Mark Teixeira
10. Matt Holliday

Here’s where some of the other guys rank:
Mauer, 16; Nelson Cruz, 20; Granderson, 29; Zobrist, 37; Uggla, 65 (but still 5th best 2b); Dunn 50; Berkman, 42; Ross, 51; Gutierrez, 70; Davis, 26 (that’s a big difference…); Rios, 58; Span, 88; Roberts, 110; and Ichiro 17 (I readily admit that I messed up on him).

I think this formula works for me. But it sounds like it’s been vetted.  Obviously, I still need to make some kind of correction for position, though.