Mythbuster #5: “Veteran teams are better equipped for tourney success.”

We’re about 50 hours away from Selection Sunday. So that means we’re about 50 hours and one minute away from some college basketball pundit spouting a completely unsupported axiom about what it takes to make a deep tourney run. You’ll hear all sorts of B.S. (hey! aren’t those the initials for Bracket Science?) over the next week. And we’ve created this Mythbuster series to separate the facts from the fiction.

Here’s one argument some former coach is sure to advance: “Well, Jay, they’ve got a senior-laden squad, they know their system inside and out, so you know they’ll be dangerous.” There’s an assumption in a lot of sports that veterans have a natural advantage of youngsters. Is it true in the tourney?

I looked at team age in two ways: 1) by the number of seniors in the starting lineup, and 2) by the average class of the starting unit, where a freshman equals one, a sophomore equals two and so on. Then I did a standard PASE analysis of one through six seeds over three nine-year periods—1987 to 1995, 1996 to 2004 and 2005 to 2013. (I don’t have team age data for 1985 and 1986). The reason I stuck with the top six seeds is that they’re the likely deep advancers in the tourney. And I broke the analysis into three eras because, with the one-and-done rule and other changes in college basketball, there could be a trend in the team age dynamic.

In fact, for both the analyses I did, there is a trend—and it’s decidedly against older teams. Let’s look first at the PASE performance comparing teams with different numbers of senior starters:

myth_exp_1

Teams with more than two senior starters have performed progressively worse against seed expectations over time. In the earliest era, they were slight +.077 PASE overperformers. Then they dropped in the middle years to -.148 PASE underachievement. And in recent years, they’ve fallen short of seed expectations by nearly half a game per tourney. Meanwhile, the teams with no senior starters have always been overachievers, as have those with one senior starter. Since 2005, however, the senior-less teams are clearly the best performers of the four groups.

The story is basically the same when we run the numbers based on the average age of starters. The youngest teams (averaging less than 2.8 on the 4-point scale) were narrow underachievers way back between 1987 and 1995. But since then, they’ve been the biggest overachievers of the three groups. Meanwhile, the oldest teams (greater than 3.0—or all juniors—on the 4-point scale), were decent overperformers early, but have been getting gradually worse over time. Check it out:

myth_exp_2

Interestingly, the youngest teams used to be the only age group that didn’t meet seed expectations. Now, they’re the only age group that exceeds them.

The Verdict: Do not listen to any expert claiming that a squad of grizzled veterans is better prepared for a deep tourney run. The evidence doesn’t support it. Let’s face it: in this one-and-done era, the teams with older starters are less likely to have star talent. Otherwise, that talent would’ve already jumped to the NBA.

This entry was posted in Mythbuster. Bookmark the permalink.

6 Responses to Mythbuster #5: “Veteran teams are better equipped for tourney success.”

  1. Ryan says:

    Last year’s Miami team is a perfect example.

  2. Jared says:

    However, you have to look at last year’s Kentucky team that did not even make the tournament, thus somewhat skewing the results. They had the most talented young team in the nation and went out in the 1st round of the NIT, so they cannot even be represented by the horrible PASE that they would’ve shown (I understand that it would not have been that significant of a difference but still). Also I would like to see how older mid-major schools compared to older big 6 schools and how younger mid-majors compared to young big 6 schools. My guess would be that older mid-major schools are better for pulling a cinderella run because of some of the reasons that these supposed “analysts” say, however it would make sense that senior laden big 6 schools do not do as well as they should because of the reasons stated above. But splitting this category into mid-major and big 6 would be very interesting to see.

    • ptiernan says:

      There’s something to your point Jared. I did a quick analysis on Big Six and non-Powers after 2005…and I further divided into 1-6 and 7-12 seeds. For teams older on average than juniors:

      1-6 Power: -.077 PASE
      1-6 NonPower: -.365 PASE
      7-12 Power: +.043
      7-12 NonPower: +192

      It may be correct to conclude from this that older teams overachieve at lower Cinderella seed positions, but underachieve as frontrunners.

      Here’s the same analysis for the youngest group of teams (<2.8 out of 4.0)

      1-6 Power: +.076
      1-6 NonPower: -.018 PASE
      7-12 Power: -.108 PASE
      7-12 NonPower: -.351 PASE

      No doubt at all about this: veteran underdog-seeded nonPowers are definitely overachievers while their younger counterparts are massive underachievers. That dynamic is not at play for the higher seeded nonPowers.

      Good contribution, Jared!

  3. Mike L. says:

    Maybe MICH with all their youngsters will bring the YOUNG PASE crashing down with a 1st rd loss this year.

    Just picking with you, Pete.

    On a more serious note, the ranges of the groups in both charts are noticeably different. In the 1st chart, the ranges of the 0SR and 1SR groups are 0.052 and 0.094 respectively, but the ranges of the 2SR and 2+SR groups are 0.209 and 0.533, respectively. Likewise, in the 2nd chart, the ranges of the YOUNG, MIDDLE and OLD groups are 0.086, 0.158, and 0.189, respectivley. It seems as though the fall in the older groups isn’t met with an equal and opposite rise by the younger groups. If I had a guess, I would say that the fall in the OLD group may be due more to the changing age-profile of the college game (teams getting younger and younger) than anything statistically measurable.

    I’m also curious about your selection of the Groupings in the 2nd chart. YOUNG was defined as 3.0, which leaves MIDDLE as 2.8-3.0, not as wide of a margin as the other two groups (1.0-2.8 & 3.0-4.0). Also, you don’t have to do this because I know you are busy, but I think a minute-weighted grade of team age may be more telling. This kind of study has so many factors at play, it may be impossible to get an reliable or valid measure of age’s importance to tourney performance. A great thought-provoking article nonetheless and that’s why I love your B.S. website (that is Bracket Science website, LOL). (Note: All credit for this pun goes to the original author, Pete). Again, great work.

    Mike L.

    • ptiernan says:

      Mike – the disparity in the decreases and increases can be attributed somewhat to the numbers in each group. There were fewer younger teams in the early years and more in the later era–vice versa for older teams.

      Hey…I’m not sold on Michigan this year. I think it’s a risky proposition to have to outshoot poor defense six games in a row.

Leave a Reply

Your email address will not be published. Required fields are marked *

* Copy This Password *

* Type Or Paste Password Here *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>