Preseason polls are just that—preseason something to look at and talk about. The real work is about to begin!
Only 8 SEC teams in the top 15? No Miami/SWAC, and they're playing with my emotions. I'd sign up for that if it were a forecast. Go GATORS! ,WESGATORS
No Bama was surprising to me(top 10 ish IMO),, or any of the Miss schools.. S. Miss and those other 2. 8 in top 15,,but none after was a bit strange.
All media- paper, broadcast, digital- exist to fill the space between the ads. Publishers like stuff that creates controversy because it makes readers/viewers/listeners come back again. Not because they'll read more areticles, but because they'll see more ads. If the substance of article is strong accurate news or analysis, that's nice, but some people come back to see more garbage content if garbage appeals to their individual biases or prurient interests. Even better if it includes names and faces for people you know or wish you knew. That approach made billionares of Rupert Murdoch, Mark Zuckerburger and others.
I know voting should not be this way but I guess if teams start out highly ranked and win enough games then it vindicates the voters for putting them up there in the first place. Which means the voters resist dropping them down for teams with better records. Yeah I know sounds like a bunch of crap. But it means its nice to start higher than most. That said its far more important to get 20 regular season wins in the SEC to assure a team is in the top 8 and possibly a conference champ.
In fairness to the pollsters the 2 Miss schools in the sec were terrible last year, tho I'm guessing they can't both be as bad this year. Bama was better but has a new coach and no baseball pedigree to get votes. I see Arky rated high again, maybe one of these years they will get over the hump. Hopefully not this year.
Ken Pomeroy found that in college basketball the preseason AP poll is actually a better predictor of post-season performance than the last AP poll of the year. He reasons that this is because pollsters put thought into their preseason polls based on many factors, but during the season their adjustments are often reduced to the overly simple formula of “win = stay; lose = drop”.
I'm not disagreeing but basketball has different conditions than baseball such as one game vs 3 day weekends, indoor environment vs possible light rain and wind blowing out to right, hard wood courts vs grass and dirt. I could go on but I'm sure everyone gets it. Your point is interesting that a preseason poll might be a better predictor than the experts having actual results to go on. Crazy I get it. I'm just maintaining how difficult baseball can be and how important conference league play and weekend series can affect the games. I do think the game factors that go into a baseball series or a season matter more to the win loss success in the polls than the game factors that affect basketball polls. Maybe that's why a preseason basketball poll might be easier to predict the end of season outcome than baseball. Maybe???
Your post raises two interesting questions: First, to what extent can we apply lessons from college basketball polling to baseball? I don’t know exactly. I certainly take your point that baseball has added facets of variability, such as starting pitching. Indeed, we’ve routinely seen a team win and lose by double digits to the same team in consecutive outings. Does that mean that Ken Pom’s insight about preseason basketball polls cannot be applied to baseball? Maybe, but it also might suggest the opposite, that this pattern might be more true in baseball. If the issue is that outcomes during the regular season lead pollsters away from accurately representing team strength in basketball, perhaps the added variability of baseball exacerbates this very problem? Then again, this effect is likely ameliorated to some extent by the baseball tradition of playing 3 game series instead of one-offs like basketball. Second, how much can we ever know in absolute terms about the strength of a college baseball team? Above we think we discovered that the knowledge included in a preseason can exceed that contained within a postseason poll, but this doesn’t tell us that this is any kind of respectable feat. Perhaps preseason polls predict just 6% of outcomes, but postseason polls predict an even more dismal 4%. Perhaps all polls are terrible (however defined) as has been suggested above, but do we have anything that is better? Is it possible to do better? Indeed, this has been a project of many a statistician and gambler. If someone simply doesn’t believe we can know how good a baseball team is, then it would appear that polls must be bad, but so should every other metric gauging team strength. On the other hand, if someone says that the polls are underrating team X, it would appear this person believes they’ve identified a better method to assess team quality than the media/coaches being polled. But if this true, one wonders why these other people haven’t adopted this superior method. I am personally in between. I don’t think polls are great, but I think they usually do indicate something real. I do like the quantitative algorithms better, but they aren’t perfect either.
I might not have the exactly correct numbers but I am sure that I have them in rank order. In football the better team wins 90% of the time. In basketball the better team wins 75% of the time. In baseball the better team winds 60% of the time. There is way too much randomness in baseball. That is why they need a 162 game season at the pro level.
