Current odds to win the 2015 World Series

Yes Rad, Vegas builds it's "juice" into its world series odds, whereas I posted "clean" odds that should add up to 100% or very close to it. Part of the Indians' edge in my opinion is the overall weakness of their division. From top to bottom the other 2 AL divisions are stronger. If the Tigers don't re-sign Scherzer I think it will be a VERY close race. I think the Royals are due for enough regression to push them behind the Tribe, if not by a lot.

I just look at the K rates of that young pitching staff and see the chance of Cinderella to get the glass slipper. To a much lesser extent, I think the same can be said about a team like the Mets. But even there, they play with the Nats in their division; baseball's most talented team.

A team that wins a division has a roughly 1/8th chance to win the World Series since there are 7 other teams remaining once they start postseason play. The postseason is enough of a crapshoot, and we are far enough away from Opening day anyway, that I use the 1/8 multiplier once I figure a team's odds of winning it's division. Once we approach the season's beginning I tweak, but only slightly, the 1/8th multiplier based on my assessment of whether a team is particularly well-suited to win short-series in October. But really, 1/8 is the appropriate multiplier in almost all cases.

By contrast the two wild card teams must win a play-in game just to advance to the level of the final 8. So the appropriate multiplier there is 1/16. Win a coin flip game then become one of 8 teams in a postseason tournament crapshoot to win the World Series. No it's not a pure crapshoot but it's close enough to one that the multiplier works at this early stage of the offseason.

So once you've figured out an estimate for a team to advance to the postseason as a division winner you multiply that by 1/8. You then add that to the product of the estimate of winning the wild card and 1/16. From that comes a percentage chance to win the world series, which can be expressed in Vegas style "odds against" as I described earlier in this thread. I.e., a team with a 10 percent chance to win the World Series at season's beginning has a 9:1 shot. That's the payout that the team would get if it won it all. 9x$100 = $900. The other 90% of the time you lose your $100.

That's the mechanical and relatively easy part. The hard part is making an initial prediction about how many games a team is likely to win at season's end. That's also the fun part. I have my own "system" for doing that, but my basic starting point is to look at quality prediction systems like zips, steamer, and pecota and form a sort of "average" based on how much weight I give to a particular system based on its track record. But I incorporate my subjective opinions too, I don't just wholly rely on those outside systems.

And of course predicting that the Dodgers are likely to win, say, 90 games is really predicting that is the median probability with about a 5 game error bar on either side of that prediction. Which is where the probabilities of winning divisions come from. My prediction for the Dodgers is rosier than for the Giants, but probability being what it is the Dodgers will still lose out a certain amount of time to the Giants. just like when you simulate 1000 seasons using Strat, one team may win 70% of the time, a second divisional rival may win 15%, a third 5%, etc etc.

And those 3 prediction systems all rely on "WARP" in one form or another, not batting average, error percentage, rbi, or other counting/rate stats that measure partial performance.

A team that wins a division has a roughly 1/8th chance to win the World Series since there are 7 other teams remaining once they start postseason play.

This isn't correct. It makes the wrong assumption that all division winners have equal chance to win the title, and the wrong assumption that a division winner has more chance to win than a Wild Card team. So, your 1/8th chance method is both speculative and ignores individual team elements. And just having to play an extra game does not automatically halve Wild Card teams' playoff chances.

The hard part is making an initial prediction about how many games a team is likely to win at season's end. That's also the fun part. I have my own "system" for doing that, but my basic starting point is to look at quality prediction systems like zips, steamer, and pecota and form a sort of "average" based on how much weight I give to a particular system based on its track record. But I incorporate my subjective opinions too, I don't just wholly rely on those outside systems.

This is also unclear. How exactly do you apply "zips," "steamer" and "pecota" to formulate your "estimate" rating you give each team? To actually show your system at work, it would really help if you did an analysis of an actual team. Otherwise, your system seems--and I'm not saying it is--a bit arbitrary.

The 1/8th chance IS correct over most teams, most circumstances. But see also that I deviate from that ratio slightly as team rosters become more set closer to opening day. Read everything that I said before responding. That helps.

Also, go back and look at expert predictions over several years at the beginning of postseason about who will win the world series. it is very hard to find an expert with a success rate higher than 12.5% or 1/8th. Add all the rates up and they sum up to about...you guessed it...a 12.5% success rate. So the postseason is primarily a crapshoot. Not entirely, but mostly.

The system isn't arbitrary, I weight the systems and sum their predictions to form an average win prediction per team, and tweak according to my subjective evaluation as well. I'm just not going to spend hours providing you the entire "code" to my prediction system, or programming, as it were.

The 1/8th chance IS correct over most teams, most circumstances. But see also that I deviate from that ratio slightly as team rosters become more set closer to opening day. Read everything that I said before responding. That helps.

I did read everything you said. I was addressing your current predictions, so your deviation didn't apply. Also, if you deviate from that ratio--as opening day approaches--on teams you pick to make the playoffs, but not the ones who don't, your continual application of the original ratio becomes questionable.

The system isn't arbitrary, I weight the systems and sum their predictions to form an average win prediction per team, and tweak according to my subjective evaluation as well. I'm just not going to spend hours providing you the entire "code" to my prediction system, or programming, as it were.

You didn't show how you "weight the systems and sum their predictions to form an average win prediction per team." Until you do, and you haven't yet, it will still appear arbitrary. To prove otherwise, show your system at work. If you actually believe in it, you have no reason not to do so.

And you don't have to provide me you "code." Just describe your system in detail and show how it works.