I'll second the reader's suggestion two above about adding the average 5x5 stat line and the top 20's average. Also agree that adding a tier system is likely the way to go versus a straight top 20. I'm not sure exactly what the measure is for defining the cut off points. It's usually easiest for the top tier, but can get pretty difficult the further along you go.

This review of the rankings is a stellar idea. I do the same every year for my personal rankings. While we all know there's nothing you can do about injuries, the review process is highly enlightening nonetheless.

I think ScottyB has a great point to combine the top 20 with a tier framework. The more information, the better.

Thanks for all the insights.

I would love to see the addition of $ values for auction leagues.

Marc, tiers is pretty much how I draft so I am all for that.

Really I think the problem with straight rankings is people pretty much read the comments rather than use them pro actively so if you are looking at strategy I would say perhaps it may be a good idea to compare the tiers to ADP at a site like mock draft central and comment e.g. the next tier don't take until round 8 or something of that ilk.

You could put in a rough guide next to them e.g. 4th round (37th-48th overall)

I agree completely with including tier levels, but numeric ranking is still important.

I would suggest that you include more direct reference to players at their secondary positions, particularly if they have equal or more value there.

For instance, last season if you felt that Miguel Cabrera equally valuable for fantasy at 3B as 1B, then I think he should be mentioned in that discussion of the 3B players. If you felt that Miggy would be a five-star player at 3B, then you can just list him at the end of the five-star players to indicate he's eligible at the position. Same for Youkilis. You could leave these secondary position playes out of the numeric rankings for the 'purity' of the position ranks, and leave a complete discussion on them under their primary positions.

Three cheers for tiers! Agree that's the way to go. It's particularly useful when you're trying to settle on overall rankings.

Agree that traditonal stats, though not as informative about a player's abilities, are the reality in the majority of fantasy/roto leagues. As such, they are useful--no, vital--to each person's fantasy rankings.

Tiers.

Mark, another miss (no offense) was Russell Branyan. I don't know how you account for 30-something year old sluggers who have never played full time, but he was named the full-time 1B before the season began, and many of us targeted him for the same reason we targeted Morales - the upside, regardless of age. May need to add a "diamond in the rough" category to predict/suggest the Branyans of the world.

I was going to add this to my previous post, but it's a different topic. Keeper leagues are more and more prevalent now than ever. I don't play in any leagues that AREN'T keeper leagues, one of which is 20 team, 20 players, with 10 kept per year and a prospect drafting system. I know you don't care about the details, but working from a player pool that is minus its top 400 players makes prospecting and keeper-sleepers pretty important.

My point is, having a column for "2010 value" and a column for "long-term value" would be a really great addition that I have not been able to find at any other site.

On a long-term list, Justin Smoak could rank top-10, but he might not sniff it on a 2010 list, and therefore the list gets much less valuable for us keeper-leaguers.

(For all you nit-pickers, I'm not suggesting Smoak is a top-10 1B yet, he was just an example.)

Incidentally, I expect a big rebount for Chris Davis next year. Clint Hurdle will be a big help to him (in ways that Rudy J hindered him), and he had a big second half after his repromotion. Plus, he's kinda playing for a job in 2011, as Smoak will see the majors next year and be pushing to play fulltime in 2011.

marc, well done overall.

When looking at the players, i'd like to see some sort of risk score. When i see how Davis, for example, turned into a real bust, I think the risk warning would be appropriate. At the outset of the season, pretty much everyone was aware that Davis had enough power to hit 35+ bombs, but could also easily top 200 strikeouts for the season. As a younger player, there's always the risk that he starts off in a slump and gets benched, platooned, sent down, or at least moved down in the order. On the flip side his upside would be something more like what Nelson Cruz ended up doing.

There's something to be said in favor of, say, Justin Morneau, if he starts the season in a slump, he's going to get the benefit of the doubt and he'll stay in the cleanup spot and will at least get some counting numbers. Even though Davis may project for a PECOTA line comparable in value to Morneau's, the fact that Morneau has more of a guarantee of playing time.

Injuries aside, I wouldn't mind seeing an upside/downside figure on ABs for hitters. This is particularly handy when thinking about outfielders - it can be tempting to snag a good hitter like Hideki Matsui or JD Drew and forget that they might be limited to 400 ABs because they get platooned. In a lot of leagues (and of course you have to adjust for format) i might prefer to just take a David DeJesus, who i think might get more playing time, and provide similar value at a lower price:

Here's 5x5 stats for those 3 this year:
Drew 84 24 68 2 .279
Matsui 62 28 90 0 .274
DeJesus 74 13 71 4 .281

Thanks!

"Regular" fantasy stats, and more than 20 players.

I'm in a deep AL only league - the fun starts in the players ranked 20-30 at their positions.

First off, great job as usual. You weren't the only one that swung and missed on Chris Davis.

I think ranking numerically is fine, with tiers inserted as I think most of us do rank players in that manner. Projected standard fantasy stats would be fine also, however, I think that is available in the player projection spreadsheet produced annually. I also like that BP utilizes non-standard stats, as I find that if players deliver on non-standard stats (OPS, Total Bases), the counting stats are there at the end of the year.

It would be nice to expand with a sentence or two on the "just missed" as all of those players mentioned above are drafted in standard 12 team, 5x5 leagues and sometimes this are the best value players of the draft. Also adding suggested auction values for standard 12 team, 5x5, $260 budget would be helpful.

Another vote for tiers.

And another vote for a risk score. In my opinion, this is more important than the ranking and the one area where PECOTA has a clear advantage over the other projections out there.

The difference between winners and losers in fantasy baseball is the management of risk. Every experienced and informed fantasy player comes to the draft/auction table with roughly the same mean expectations about how players will perform in the upcoming year. The ones who win are the ones who have their first round/high $ picks perform to expectation and grab late round picks who outperform expectations.

And, although it may be heretical to say this here, PECOTA's weighted mean projections are only marginally more accurate than most of the sabrmetrically-based projections out there on the web. Where PECOTA really stands apart from the pack is the fact that the PECOTA projections are a distribution, not just a mean projection. When the PECOTA weighted mean spreadsheet comes out each year, I ignore the projected numbers and go straight to the "Break, Improve, Collapse, Attrition" columns. That tells me which major hitters to target with my early round picks (i.e., those with a low Collapse rate) and who to target with my late round picks (i.e., those with a big Breakout rate). So, these things would be great to add as well.

Marc, thanks for your analysis! I enjoy reading it every time. You do a great job!

The only adjustments I made with your 2009 forcast were that I ranked Morneau and Youkilis 9th and 10th respectively ahead of Pena. After having Davis on one of my 2008 rosters when he got called up, I wasn't sold on him. He struck out way too much with a lack of plate discipline. Like you, I ranked Votto high and was able to grab him in drafts after players like Morneau, Youkilis, and Lee were taken. He made a great utility guy for me. Good call on Morales too. I was able to get him on 2/3 of my rosters last year.

Here are some thoughts:

1. Tiers are the way to go. ADP would be useful too.

2. When you get to the OF position, it would be very helpful if you continue to rank all 3 OF positions in conjunction with all 3 combined into 1 ranking.

3. Other stats that I think would be useful are BB/K ratio and ISO power. It's my belief that AVG/OBP/SLG are key indicators for a good ball player. Yes, most people play in standard 5x5 leagues, but when forecasting a players likelihood, stats like OBP and OPS are more useful.

4. It would be helpful that players who are eligible at multiple positions would be ranked at those respective positions.

5. Who saw Mark Reynolds (1B eligible in all of my leagues with 28 games played there) having the year he did? He was a late round flier for me since he is the king of strike outs, but wow! And he stole 24 bases! It will be interesting where he's ranked.

I would like to see a slightly longer list, as I participate in an NL only league.
Wouldn't mind seeing PECOTA and actual EQAs for players.

Just did a quick analysis of the data (dropping Jackson & Delgado).

Here are PECOTA vs. actual correlations and deltas

BA OBP SLG
Correlation 0.72 0.82 0.73
delta(PEC-actual) -0.006 -0.012 -0.025


PECOTA did better predicting OBP than BA or SLG
PECOTA on average is conservative at predicting actual values

Repost Table with spacers

....................BA... OBP... SLG
Correlation.......0.72...0.82...0.73
delta(PEC-actual)-0.006 -0.012 -0.025

I second the suggestion to expand the list beyond 20, not necessarily by ranking beyond 20 but as someone else suggested, with a "just missed" category.

I also really like the idea of highlighting collapse/breakout potential.

I'd like to see tiers and we use OBP and SLG in our league.

I'll also put in my 'vote' for a tiered system. Standard 5 X 5 scores are fine, but having triple slash stats on top of it would be helpful. As far as risk assessment goes, I am fine with the idea proposed by others, but to me, that's where Will's Team Health Reports come in.

Finally, having a 'sleepers' tier would be of great benefit. These could include guys like Kendry Morales last year (your mention of him last year re-affirmed my faith and helped me to decide to pick him up..thank you). You could also throw in a couple of prospects there (not necessarily the best prospects, but the ones most likely to have fantasy value this season). I know that there are many of you in keeper leagues (I want to get in one myself). While including long-term investment rankings is nice, I assume that anyone in a keeper league is going to be reading as much as they can on this site.

Overall, the key to a good rankings list is to keep it meaningful and concise. The more 'stuff' you add to it, the more you run the risk of making the rankings cluttered, thus decreasing its usefulness as a tool.

I agree that a tiered system, with perhaps a base ordering within the tiers, would be the way to do it next year for all positions.