2013: Are the Cubs .500?

Last year, the Cubs won 61 games. In the offseason, Chicago made the following signings:

Edwin Jackson

Scott Baker

Carlos Villanueva

Scott Feldman

Nate Schierholtz

Scott Hairston

Dioner Navarro

Kyuji Fujikawa

The rest of the team is more-or-less intact: the Cubs basically had no free-agents (they traded everyone that was going to be a free agent of worth).

There aren't 20 wins worth of additions there… so why do I have the Cubs predicted to win 81 games?

The Spreadsheet

2012 PA     PA WAR 12 WAR Note  
630 C Castillo 400 1.2 0.6    
    Navarro 230 0.5 1.1    
680 1B Rizzo 590 3.7 1.6    
    Hairston 90 0 2.1    
642 2B Barney 580 2.9 1.8    
    Valbuena 60 0.1 1.2    
698 SS Castro 680 4.3 3.3    
    Valbuena 20 0 1.0    
647 3B  Stewart 400 1 -0.2    
    Valbuena 250 0.9 2.1    
682 LF Soriano 500 1.1 2.3    
    Valbuena 180 0.2 -1.0    
672 CF DeJesus 450 1.7 0.5    
    Sappelt 220 -0.1 1.1 AAA Tr*  
693 RF Schierholtz 400 0.9 2.2    
    Hairston 300 0.8 -0.5    
  P All of them 320 -2 7.2 2012 Delta  
  PH All of them 270 1.7 1.9 Repl Adj  
  Team PA Total 5940 18.9 5.2 2012 Delta**
    PA/G 36.7        
33   Equ. DL G 150        
    % Repl PA 10.3% 17.0      
               
               
* MLE from PCL numbers last year.          
** Difference from last year, adjusted for % replacement PA.    
    IP WAR 12 WAR Note
SP1 Samardzija 169 3.1 9.2  
SP2 Jackson 187 2.9    
SP3 Garza 151 2.8    
SP4 Feldman 109 1.3    
SP5 Baker 111 1.8    
SP6 Villanueva 111 1.7    
SP7 Wood 102 0.9    
RPSP None 0 0 14.1  
Bullpen Everyone 500 -0.4 -1.5 4.35***
      14.1 1.1  
        15.2  
        2.1  
IP Targets Starter 940   13.1 18.3
  Reliever 500      
           
           
           
           
           
           
           
  % Repl IP 15% 12.0    
        TWAR 28.9
        REPL 51.8
          80.8
           
*** 4.35 totalRP        

This will take a LOT of explaining, so I'll start from the beginning. 

My first step was to assign playing time to each person at each position. For instance, Castillo is the starting C, but even in a perfect world he's only going to play 2/3 of the time. I "ideally" gave him 400 PA and Navarro 230 PA (the Cubs had approximately 630 PA come from the catchers last year). I did this for every position: these numbers are what I would predict in each case if there are no injuries. I'll explain injury considerations later.

After I've put a reasonable approximation for each players' PT by position (assuming that Stewart is the primary 3B and Hairston is the primary backup 1B – my projection doesn't assign a single PA to Clevenger), I then tried to project how much WAR they would provide given their plate appearances, defensive skills, and baserunning abilities. I used this handy calculator to do so. In the spreadsheet, I also made a note of how much value the Cubs' received from each position last year, as well as the difference. In most cases, there is a significant upgrade; in fact, only 2 positions regressed, the corner outfields. This makes some sense to me, though I very much expect the gains to be overstated in these projections. To avoid as much bias as possible, I used every players' ZiPS projections exactly. A short summary of each position follows:

C: Soto was really bad last year, and Clevenger was worse. I'm pretty high on Castillo, but you don't have to be to have a marked improvement here.

1B: A full year of Rizzo will be the most important and maybe the most difficult thing to project. I think 3.7 WAR is optimistic here. Important to note that Hairston is a replacement-level 1B.

2B: I'm pretty low on Barney. His defense last year was insane and it seems unsustainable to me. It's important to note that Barney's WAR last year was higher than 1.9, but the subs for him were universally terrible last year.

SS: Castro is projected to take another step forward. I actually agree, and would not be surprised in the slightest if Castro exceeds even this lofty projection.

3B: Ian Stewart is beloved by ZiPS for some reason. I don't get it. Valbuena is the better player of the two in the system, but from a utility standpoint it makes way more sense for Valbuena to be a super sub. If the Cubs call up Watkins to fill that role, than Valbuena should be the 3B, Stewart the backup 3B/1B. I don't like the retardation of Watkins' development, though.

LF: Soriano is goign to fall of a cliff this year. We should trade this guy, according to ZiPS. 

CF: DeJesus' bat plays better in center, but he's going to really struggle there defensively (at least he did last year IMO). Our OF defense is going to be pretty bad in general, I think. I ignored Sappelt's MLB numbers completely, and instead performed a MLE of his 2012 AAA season. 

RF: The Schierholtz/Hairston platoon isn't that exciting, partially because platoons rarely work in practice. That's why I'm comfortable using ZiPS projections that don't take platoons in mind (and don't adjust for platoon advantages). Both Nate and Scott are just guys, below average but above replacement.

P/PH: I just went with the same numbers as last year. 

After I totalled that WAR (18.9 WAR), I built in a contingency for injuries. The Cubs have lost, an average 150 total games to injuries over the last 3 years. That's 10.3% of their roster given to replacement-level players. One of my primary assumptions for this projection is that the injuries affect a team at random; no player is more or less prone to injury as another. I know this isn't the case, but I'm also sure that I (or anyone else) can predict individual player injuries with a high degree of certainty. Therefore, I just "replaced" 10.3% of the team's PA with a replacement-level roster, reducing the overall batting WAR to 17. Last year, the Cubs produced 11.7 WAR, so this is a significant difference (5.3 WAR). I think it's reconciled partly by replacing players like Joe Mather (-2.1 WAR last year) and Josh Vitters (-1.3) and Steve Clevenger (-1.0) with (hopefully) replacement-level guys. The Cubs left a lot of WAR on the table last year by virtue of starting these sub-replacement players. 

For pitching, I had a more arduous task. I first had to project innings counts for each of the 7 starters. I'm not at all confident in my list, which in the end was just the ZiPS projections for the first 6 and the remaining starts to Wood. ZiPS is also unreasonably confident in the ability of everyone to take a step forward, either from injury or ineffectiveness. The bullpen is very difficult to project individually (every individual RP's volatility is high), so like our banking system I just took a bunch of volatile pitchers and lumped them together and sold the RP off as a 4.35 ERA, 500 IP monstrosity. That's a slight improvement over last year, which seems fair.

After I totalled the Cubs' WAR from pitching (14.1), I did a similar correction for injuries. I used 15% here, which is admittedly a little low: I think the Cubs are more injury prone than the average team (probably around 21 or 22%), but they also have an above-replacement pool of pitching to draw from (for SP). I could be talked down here for certain. I ended up with 12 WAR pitching, and the Cubs had 7.7 last year. Volstad (-1.9), Germano (-1.5) and Dolis (-1.2) are all that's needed to reconcile with that.

Take those 29 WAR and add 51.8 replacement level wins and you have an 80.8 win team!

Let's think hard about this.

The first thing to recognize is that the WAR from last year and this year don't match up. The difference from total WAR this year and last is around 10 WAR, but this team is 20 wins better. Part of is explained by the Cubs being "unlucky" last year, underperforming their W/L by 4 last year. The other part is somewhat explained by my using FanGraphs WAR for last year's WAR, which has a different (and lower) replacement threshold. It somewhat overvalues last year's production, where as (to the best of my knowledge) the WAR calculator I used uses the same .320 percentage threshold that B-R does. If this discrepancy is in fact significant, I can see downgrading my projection as many as 4 to 6 wins.

Next, lets remember that my projection doesn't take into account a single trade. The Cubs are likely not making the playoffs this year, and parts WILL be traded off in that case. However, my projections have Garza pitching 151 innings for the Cubs, Soriano grabbing 500 AB, and all of our flippable pitchers more-or-less staying on the team for a full season. In reality, this is likely not the case. Trading Soriano and Garza immediately docks the Cubs 4 wins.

Also, keep in mind this is a relatively healthy projection, too. The injury contingency deals with the team on a broadscale, but it's more likely than not that one major component of our team will be lost to injury. Soriano, Garza, and Baker are all huge injury risks. Losing them is a pretty big deal. 

Last, let's keep in mind that ZiPS is pretty optimistic concerning the Cubs. Is Stewart going to produce 1 WAR in 400 PA? Probably not. Is Rizzo going to produce 3.7 WAR? Maybe, but that seems high. Our the Cubs going to get average to above-average production from 7 simultaneous time-share starters? Absolutely not. Will those 7 pitchers account for all 162 starts, as I have projected (save the injury contingency). No. 

It's also not like I'm saying the Cubs are even a good team. This projection says that the in the optimal production case with average injuries, the Cubs aren't even above .500. That doesn't seem like its too far from the truth.

If I were a betting man and wanted to be a little more subjective, I'd adjust this downwards to maybe 75 wins. As it stands, though, on paper, the Cubs appear to be right around…average.

This is a definite work in progress, so please let me know what looks good, bad, and ugly.

Quantcast