Tomorrow’s Cooperstown election results today

Every year the annual BBWAA vote for Cooperstown transfixes me. I love guessing how the vote will shake out.

Over the years, I’ve discovered 10 rules of thumb that make the annual ballot quite a bit more predictable. None are perfect, but they work.

Guideline No. 1: consistency

Duh. Voters revise their thoughts, but players rarely experience inexplicable shifts in their support. For example, in the last dozen elections, Dave Concepcion has always been between 11 and 17 percent.

However, some things do cause ebbs and flows in a player’s balloting strength, hence the other guidelines.

Guideline No. 2: strength of ballot

This might be the most important rule. You got two kinds of candidates on the ballot: first-timers and the backlog.

When there’s an unusually strong crop of newbies, the members of the backlog lose support. Should there be a rather weak batch of newcomers, the holdovers will rise up. Pretty simple, but the results can be dramatic.

Let’s look at 1999. Headed by Nolan Ryan, George Brett, Robin Yount, Carlton Fisk, and Dale Murphy, it had the strongest new candidates in memory. Look at what happened to the holdovers from the ’98 ballot:

Name             1998    1999
Tony Perez       67.90%  60.80%
Jim Rice         42.90%  29.40%
Gary Carter      42.30%  33.80%
Steve Garvey     41.20%  30.20%
Bruce Sutter     31.10%  24.30%
Tommy John       27.30%  18.70%
Jim Kaat         27.30%  20.10%
Dave Parker      24.50%  16.10%
Bert Blyleven    17.50%  14.10%
Davey Concepcion 16.90%  11.90%
Minnie Minoso    16.10%  14.70%
Luis Tiant       13.10%  10.70%
Keith Hernandez  10.80%  6.80%
Dwight Evans     10.40%  3.60%
Mickey Lolich    8.20%   5.20%
Ron Guidry       7.80%   6.20%
Bob Boone        5.50%   5.40%

Altogether, the backlog lost about 25 percent of its support. Many, including Jim Rice, Bruce Sutter, Gary Carter, and Bert Blyleven, had their worst showings ever. Sutter came within 0.5 percent of his worst mark.

Well, that was the greatest ballot ever. What happens when the newbies are strong, but not that insanely great?

Last year, Cal Ripken Jr. and Tony Gwynn, were a historically strong one-two combination. However, only one other person, Mark McGwire, received over six percent of the vote, and he only polled as well as Dale Murphy did in 1999. ‘Twas a strong group, but not 1999.

Only two of the 15 backloggers—Concepcion and Goose Gossage—had their votes go up. Concepcion went up 1.1 percent, and Gossage rose up for reasons I’ll explain in a bit. Even including Goose and Concepcion, the backlog lost 10% of their support.

Alternately, a weak bunch of newcomers causes the backlog to rise up.

Guideline No. 3: comparable candidates

The best example of how comparable candidates can impact backloggers comes from the sad story of Jim Bunning, Luis Tiant, and Mickey Lolich: three big-winning, long-lasting starting pitchers.

In 1988, Bunning scored 74.2 percent of the vote in his 12th year on the ballot, painfully close to entry. Tiant appeared on nearly a third of the ballots and Lolich a quarter.

Then disaster struck: next year Gaylord Perry, Fergie Jenkins, and Jim Kaat showed up. Even worse: because Johnny Bench and Carl Yastrzemski appeared, none of the new pitchers got elected.

In 1990 Jim Palmer won election on arrival, while the others stayed in the backlog. In 1991, Jenkins and Perry finally went in. Then came the steady stream of 300 game winners. Tom Seaver, Phil Niekro, Steve Carlton, Don Sutton, and Nolan Ryan all took turns debuting. At least one 300 game winner was on the ballot every year from 1989-99.

Here’s the percentage the trio received every year, showing how comparable impacted them:

Pitcher	1988	1989	1990	1991	1992	1993
Bunning	74	63	58	64		
Tiant	31	11	10	7	12	15
Lolich	26	11	6	7	11	10

Pitcher	1994	1995	1996	1997	1998	1999
Tiant	9	10	14	11	13	11
Lolich	5	6	7	7	8	5

Yowch. These numbers tell a story. The writers so wanted to get Bunning in. No one wants to see someone get that close but pulled back. Heck, he even beat Fergie Jenkins in 1989. But too many pitchers outclassed him, and he couldn’t get back near 75 percent. Meanwhile, few writers wanted to put four to six pitchers on their ballot, so Tiant and Lolich had their support completely crumble. Catfish Hunter is damn lucky he went in before this litany of big winners showed up.

It’s worth noting that comparables don’t have to play the same position. Sluggers are comparable, even if one is a first baseman and another an outfielder.

Guideline No. 4: “Over the top!” surge

When a candidate gets close, there’s a tendency to surge him over. BBWAA ballotters are especially likely to look at re-evaluate those near the top of the backlog, and are willing to be talked into supporting a player’s candidacy. The best way to explain is with some examples.

Ryne Sandberg:
2004: 61.1%
2005: 76.2%

Tony Perez:
1999: 60.8%
2000: 77.2%

Bruce Sutter:
2005: 66.7%
2006: 76.9%

All candidates received some of the biggest boosts of anyone on the ballot the year of their election. Mind you, it should be harder for someone at 60 percent than 40 percent to receive 100 additional votes. There’s fewer voters left to persuade. Sandberg and Perez had to convince almost half of those who left them off to add them on. But they did.

Also, as I’ll show later on, these surges are particularly strong when the ballot has a weak crop of newbies.

It doesn’t always take only one year. Gary Carter surged from 2002-3, but had to wait until 2004 to go in. Once a surge starts, it’s hard to shut down. Jim Bunning proved it can happen, but it took a helluva lot in his case.

Guideline No. 5: relief pitcher wackiness

Relief pitchers are the most fluid creatures on the BBWAA ballot. After 70 years of voting, there’s some sort of consensus on what makes a Hall of Fame first baseman or shortstop. Relief pitchers? We’ve only got four in there, half of whom went in the last four years. The most impressive rises on the ballot in the last two decades are Sutter and Gossage, both relievers.

This murkiness on what constitutes a Hall of Fame closer is why Gossage rose up last year when almost no one else did. Sutter’s induction gives people a better idea of what the marker should be, and makes it more obvious Gossage belongs.

Something similar could happen to McGwire, as the BBWAA mulls over how to handle players believed to have used steroids. (Also, I vaguely recall some sportswriters say they’d abstain from voting on him for one year. This is year two).

Guideline No. 6: last year on the ballot

The BBWAA doesn’t let a candidate stick around forever. Eventually, they have one last shot for election. (Usually it’s 20 years after their final season, but for a handful of players like Ken Boyer and Ron Santo that wasn’t the case for whatever reason) In the last 20 years, 21 survived until their final year.

Altogether, they appeared on 22.9% of ballots in their next-to-last year, and 25.9% in their last year. Fourteen had their totals improve, one stayed in neutral, and six got worse. Most whose votes dropped were on rough ballot, such as Mickey Lolich and Minnie Minoso, who both appeared on the 1999 ballot of doom.

In 2008, Concepcion faces his last ballot.

Guideline No. 7: candidates per ballot

One thing to keep in mind when predicting how the BBWAA will do this year is how many names each ballot averages. The Hall of Fame allows a BBWAA member to pick up to 10 names on his ballot. A large majority of BBWAA voters pick less than 10. Here are the averages in the last 20 years:

Year	Avg
1988	6.61
1989	6.75
1990	6.87
1991	6.65
1992	6.07
1993	5.76
1994	6.32
1995	6.15
1996	5.72
1997	5.59
1998	5.41
1999	6.74
2000	5.64
2001	6.33
2002	5.95
2003	6.60
2004	6.55
2005	6.32
2006	5.64
2007	6.58

Historically, this is pretty low. In fact, 1987 was the first time the BBWAA ever went under seven names per ballot. They’ve never gone over since then. The worst years predictably corresponded with weak incoming classes.

Wait a second—if the total is nowhere near 10/ballot, then how come a strong bunch of newbies hurts the backlog? Two reasons: 1) some do fill out their ballot to 10, and don’t have room for more, and 2) more importantly, some voters just ain’t comfortable listing too many names.

To really sort it out, let me add a few columns: newbies per ballot, the major newbies, backlog per ballot, and those elected in the year:

Year    Avg     Newbies Major Newbies                      Backlog       Elected
1988    6.61    1.37    Stargell, Tiant, Lyle              5.24          Stargell
1989    6.75    3.35    Yaz, Bench, Jenkins, Perry, Kaat   3.4           Bench, Yaz
1990    6.87    1.79    Palmer, Morgan                     5.08          Palmer, Morgan
1991    6.65    1.72    Carew, Fingers                     4.93          Carew, Perry, Jenkins
1992    6.07    1.76    Seaver, Perez                      4.31          Seaver, Fingers
1993    5.76    2.09    Jackson, Niekro, Garvey            3.67          Jackson
1994    6.32    2.06    Carlton, Sutton, Sutter            4.26          Carlton
1995    6.15    1.54    Schmidt, Rice, John                4.61          Schmidt
1996    5.72    0.29    Boone, Lynn                        5.43
1997    5.59    0.3     Parker, Dw. Evans                  5.29          Niekro
1998    5.41    0.65    Carter, Blyleven                   4.76          Sutton
1999    6.74    3.62    Ryan, Yount, Brett, Fisk, Murphy   3.12          Ryan, Brett, Yount
2000    5.64    0.68    Gossage, Morris                    4.96          Fisk, Perez
2001    6.33    2.12    Winfield, Puckett, Mattingley      4.21          Puckett, Winfield
2002    5.95    1.54    O. Smith, Dawson, Trammell         4.41          Smith
2003    6.6     1.86    Murray, Sandberg, L. Smith         4.74          Murray, Carter
2004    6.55    1.8     Molitor, Eckersley                 4.75          Molitor, Eckersley
2005    6.32    1.03    Boggs                              5.29          Boggs, Sandberg
2006    5.64    0.32    Hershier, Belle                    5.32          Sutter
2007    6.58    2.32    Ripken, Gwynn, McGwire             4.26          Ripken, Gwynn
Avg     6.21    1.61                                       4.6

Couple points: First, generally, the worse the crop of first timers, the fewer names per ballot. Second, several years of bad newbies, such as was the case in the late 1990s, causes record lows for names/ballot. That’s key because we’ve had sparse crops two of the last three years and will have another this year. 2008 could set the record for fewest names per ballot.

Third, backloggers usually can only get in when there isn’t much of a rookie crop. While only 10 backloggers have been elected in the last 20 years, the five times less than 1.0 newbies per ballot were named, four get elected. That included 2000, one of only two years the BBWAA elected a pair of backloggers .

Finally, remember that surge stuff mentioned above? Sutter surged in one of the weakest crops of newbies ever, with Perez in one only slightly stronger. Sandberg went in with newbie Wade Boggs, but aside from him the crop was astonishing weak, with a non-Boggs newbie appearing on barely one-tenth of all ballots.

Guideline No. 8: Repoz and Keith Law

What’s a Repoz? He’s the chief article linker at baseball primer. For the last few years, he’s tallied every BBWAA who publishes his ballot . It ain’t perfect but it works pretty well. In 2005, for example, his tally was within three percent on six of the top ten.

Keith Law’s doing something similar this year at his blog. It’s a bit more extensive as some voters he knows who aren’t publishing their ballots have told Law how they voted. An element of bias creeps in as the people friendly with Law are more likely to be friendly to sabermetric darlings. In fact, in Law’s most recent update, Blyleven led Rice. That won’t be how it shakes out.

In particular, the surveys gives an idea how newbies fare. Last time I saw, Repoz, based on 44 ballots, found 50 percent support for Tim Raines, a higher total than I would’ve guessed. With 88 ballots, Law has him at 38% percent

Guideline No. 9: beware five percent

For the last however many years, 5% is the third rail. Get below it, and you’re dropped off the ballot.

This guideline means that players who get near the five percent marker almost always drop below it. The writers usually aren’t too willing to support a lost cause. Lolich is perhaps the only man in the last two decades to fall below seven percent and survive to his last ballot.

Guideline No.10: guidelines ain’t laws

Let’s not get carried away with this stuff.

What it means for 2008

The best newbies for this year are Raines, Brady Anderson, David Justice, and Chuck Finley. Raines will possibly be the only one to clear five percent. It’s an unusually weak crop that will finish short of one newbie per ballot.

Fourteen men return. Two, Gossage and Rice, topped 60 percent last time: 71.2 percent and 63.5 percent respectively. The only times in recent years two backloggers got elected came when both topped 60 percent the year before. In 1990, Perry was at 72.1 percent and Jenkins at 66.7 percent. In 1999, Fisk and Perez finished at 66.4 percent and 60.8 percent.

I can’t imagine a scenario in which Gossage doesn’t go in. All arrows are pointing up on him. He had a surge last year that combined with relief pitcher wackiness to overcome an usually strong bunch of newbies. It’s an incredibly weak ballot this year with no strong comparable players arriving.

The tricky one is with Jim Rice. I figure he’ll be within two percent either way, 73-77 percent. 2008 looks a lot like 2000 to me. You have two leading candidates on top the ballot, a weak bunch of new candidates, and an overall names/ballot total that should be very low.

Raines is comparable comparable to Rice in that he’s an outfielder, but he’s a very different kind of outfielder. Rice was an RBI man and Raines a leadoff hitter. I don’t think he’ll hurt Rice’s chances.

Full prediction

Here’s how I see the 2008 vote going, with last year’s tallies throw in for comparison.

Name	        %       2007
Goose Gossage	83	71
Jim Rice	74	63
Andre Dawson	62	57
Bert Blyleven	56	48
Tim Raines	45	XX
Jack Morris	42	37
Lee Smith	35	40
Mark McGwire	32	23
Tommy John	26	23
Alan Trammell	17	13
Dave Concepcion	17	14
Dave Parker	15	11
Don Mattingley	13	10
Dale Murphy	12	9
Harold Baines	3	5
Rest, Combined	8	XX

That works out to 5.4 names per ballot, a new all-time low. Yet it still has an inductee, a second player who just might get elected, and most returnees improving. Few notes:

My rough draft had Rice sneaking in. I keep going back and forth on him It should be extremely close. Worst case scenario for him: election in 2009.

I’m guessing only Harold Baines and Lee Smith decline. I just can’t see the old designated hitter coming back, and Lee Smith’s candidacy was based on being the all-time saves leader. As that fades further into the past, his candidacy will have trouble regaining old voters and picking up new ones.

The estimates I have the least faith in are Raines and McGwire. You’re best bet for predicting them is to follow Repoz and Keith Law’s tracking of BBWAA voters. Aside from those two, I’m confident on my other predictions. I may not get a single specific percentage right, but I should be close on almost all of them.

Clip’n’save and check back when the announcements are made.