THE STAT GUY
Veteran managers outperform rookies
By BRADFORD DOOLITTLE
The Kansas City Star
The Royals appear to be doing the right thing by looking for a veteran manager. It also looks as if Terry Collins, a dark horse who has cropped up on the elusive “short list,” deserves serious consideration.
Collins has the first prerequisite for the Royals' managerial search: He's managed (and won) in the big leagues.
As a rule, do veteran managers make more of an impact?
To examine this question, we looked at the managerial records for the last 15 seasons. We looked at each new manager and marked him as a veteran manager (previous big-league experience) or a rookie manager (no big-league experience). There were 44 rookie managers and 41 veteran managers during this period.
To measure the effect of the two groups, we looked at the two seasons before the managerial change, then the two seasons after the change, including the new manager's first season with his team. Thus, new managers needed at least two seasons with their new team to qualify for the study.
The measurement we'll use is called EFF162 (effect per 162 games). It is calculated by subtracting the winning percentage for the seasons before the change from the percentage after the change and multiplying that times 162.
The veteran managers clearly were the more effective group. The teams that hired managers with experience improved their two-season win total by an aggregate of 85 games, for a 2.2 EFF162. Teams that rolled the dice on unproven managers dropped an extra 71 games in the aggregate for a -1.6 EFF162.
Twenty-six of 41 (63 percent) veteran managers had a positive effect on their new team. Only 18 of 44 (41 percent) first-timers aided improvement.
There were eight first-time managers on the list who had a negative EFF162 who were later given a second chance with another team. Four of the eight posted a positive EFF162 the second time around — Gene Lamont, Johnny Oates, Buddy Bell and Jim Riggleman.
Lamont is thought to be on the Royals' short list of candidates but has not been confirmed.
There were also six first-timers who posted a positive EFF162 and then went on to manage another team. All six of them vindicated themselves by improving their new charges as well.
Of these “sensational six,” four of them already have managing jobs — Dusty Baker, Buck Showalter, Felipe Alou and Phil Garner. Two of them don't — Collins and Kevin Kennedy.
Collins topped our study of available managers last week based on wins added with a score of 3.3 wins added per 162 games (WA162). Kennedy placed third on that list with 1.9 WA162 (see the Stat Guy blog for details). Collins has been verified as a member of the short list. Kennedy has not been mentioned as a candidate.
Another member of the short list, Art Howe, actually predates this study as a first-time manager because he made his debut as manager in 1989. But if he had qualified, Howe wouldn't have fared very well because the Astros got markedly worse the first two seasons after he took the helm.
Howe did make the study as a veteran manager — twice. In both instances, the teams he took over (the Athletics and Mets) got worse in the two-year period after his hire. Of course, by the time he left the Athletics, the squad had won more than 100 games for two straight seasons.
Howe posted a 1.4 WA162 in last week's study, good for fourth place on the list. Lamont finished out of the running with a -1.2 WA162. Collins is the only candidate on the unofficial short list to score well in both studies.
So it appears that the Royals are talking to the right candidate (Collins) and the right kind of candidates (veteran managers).
Of course, talk is just that — talk. Who actually gets the job remains to be seen.
To reach Bradford Doolittle, send e-mail to mailto:firstname.lastname@example.org
Spoils of experience
Veteran managers seem to have a more positive effect on a team's victory total than first-time managers. A look at new managers in the period from 1990 to 2003:
Category Veterans Rookies
Number of managers 41 44
Change in winning percentage +.013 -.010
Effect on wins +84.6 -70.5
Effect per 162 games +2.2 -1.6
Managers with positive effect 26 18
Managers with negative effect 15 26