interesting splits (1 Viewer)

[Bob Magaw]

i'll try & add to this as i go...

sometimes when an offer is made for a player on one of my IDP dynasty teams, it can act as a prompt to look at them more closely & in depth... a few that came to my attention recently...

KC SS bernard pollard... i forgot that in some of my scoring systems he was around top 20 DB overall, & close to top 10 among safeties... what roared off the page, though, was projecting his second half outburst over a whole season... it would have propelled him to close to #1 overall DB... i'm guessing some peolple may be unaware of this second half arc & trajectory, which could foreshadow bigger things to come this season...

of course, to be fair, if we projected his second half we would have to do same for other top 20 he was intermingled with by end of the season, & some of them may have done even better (probably not too many, though)...

MIA S/CB jason allen (default listed as CB in MFL, but i think he will be listed as S in '08) was another player that stood out, & his second half outburst almost mirrored that of pollard... he is a tremendous athlete with 1st round pedigree that languished to begin his career... part of the problem may have been bouncing from S-CB-S... safety may be his most natural position, & he seemed to find a home there last season, & for the first time in his pro career, his formidable athleticism, instincts & natural playing ability seemed to come to the forefront... he also would have finished close to the top of DBs overall, projecting his second half over a season...

another issue to be wary of in dynasty leagues, is these IDPs were on pretty bad teams (MIA & KC draft 1 & 5, respectively), with bad offenses, so the defense was on the field a lot... that may continue, at least for a while (though parcells has been a turnaround artist virtually everywhere he has been before)...

it would be interesting to do a study & see how many top 20-30 IDPs by position are on "bad" defenses (how to define that... bottom half, bottom 10 ??), not sure there is a profound connection... its just something to be aware of, & while a crude heuristic, in general, its better for a defendser to be on the field a lot than a little... certain players, like bob sanders, have the transcendant talent to overcome fact that he plays on a team with a great offense, that keeps it off the field for long stretches...

* sometimes trying to find patterns in data can lead to mischief... i think doug drinen did an article about how if you looked hard enough, you could be tricked into seeing spurious connections, like a player does better against teams with dark colored jersies... obviously ridiculous & irrelevant...

OTOH, if a player strings together a strong second half, especially if it is coupled with IDENTIFIABLE REASONS (for instance, this was why i cited allen's athleticism, talent, pedigree... i had expected him to do well, & it was a head scratcher when he didn't have an auspicious debut... i suppose you could say i was primed to recognize breakout signals "emanating" from his unfolding career... ), either physical, or due to change in starting status, imo that holds greater promise for helping to recognize & identify some kinds of underlying, latent, intrinsic value...

 

[Jene Bramel]

Awesome idea for a thread -- if we can get people to share what they find .

Examining splits is one of the ways I like to identify up-and-coming (or rebounding or declining) DEs and CBs. It's not often you find a killer trend, but when you do it's usually one that's flying under the radar for many. Check out the split stats for Derrick Burgess last year for a reminder what he can do when healthy.

I've been trying to force fit the "bad defenses generate good IDPs" corollary for a couple of seasons. Two years ago, I made the argument in a IDP>>>Team D league thread that the 20 highest tackling LBs were more likely to be the more talented players than scrubs racking up numbers. This year, I've been playing around with the stats of that raw tackle opportunity metric -- pass completions + rush attempts + sacks. Over the past four years, the number of teams who would fall outside a standard deviation of historical averages is pretty small. Only one or two per year (in either direction -- higher or lower opportunity) fall 2SD outside the historical means. And when I tried to pull out the "better than expected IDP production for talent" guys retroactively, they were pretty few and far between.

In other words, I believe that there are a whole lot more guys like Patrick Willis (who use their elite talent to take advantage of increased opportunity and weak surrounding cast) than Paris Lenon.

I have a couple of other ideas on how to play with the tackle opportunity stats, but haven't gotten there yet.

Still worth finding the Lenons, though, and I'm in general agreement with the more opportunity is better. In April 2008 terms, it'd be interesting to see which way a guy like Keith Brooking -- MLB on a defense that should get big opportunity, but is a declining talent playing out of position with a pretty solid surrounding cast -- falls on that spectrum.

 

[Rozelle]

Last season I started tracking Plays Against/Rush/Pass/Total, from this data I took the top 20/30 in total points scored from my dynasty league. I suppose this would be considered a tackle heavy league. Unfortunately I have only one year of data.

Solo 2

Assist 1

Sack 4 (+ .1 for every yard)

FF 4

FR 4 (+ .1 for every yard)

INT 4 (+ .1 for every yard)

PD 2

TD 12

See attachment below

Total_Plays_File.xls

 

[spec1alk]

I believe you could weight scores based no the playing time. The Time of Possession (TOP) is recorded for each game. Simply subtract the number from 60 and you will know how long the D was on the field.

If a LB gets 10% more tackles than a LB on another team, but he was on the field 20% longer, he is not really better than the other guy he just has more opportunity.

The only problem with this is for projecting in the future. It is hard to predict how good or bad a team is likely to be from year to year. CHI was a superbowl team last year and missed the playoffs this season. CLE was a bad team in '06 but nearly missed the playoffs this season and has a solid chance to win their division this year.

 

[Bob Magaw]

nice posts, guys...

i have some more IDPs to recommend to your attention, but first a few thoughts about above comments...

very good point about it being hard to predict how a team will do from one year to the next... that is one reason i like to have my eyes on a few potential waiver wire pickups, & hopefully by adding a few it increases the chances you get a yahtzee...

if you scoop up the right guy BEFORE they break out (& for our purposes, that could be as little as ONE season before they break out, or even IN-SEASON sometimes), you can sometimes get them cheaply (or get them period), before everybody else... or if you already have them & find their production flukey & maybe hard to replicate, at least you have ammo to parlay them into an upgrade at another position for a player trade, draft pick, etc...

i'll try & work through the positions systematically, but i think there may not necessarily be a lot of guys like pollard & allen at other positions this season...

one reason i like CBs (in the case of most recs below), is that when looking at the scoring leaders of past few years, there seems to be a fairly even distribution of CBs & safeties in top 20 DBs overall (& even SS & FS among safeties)... where in a lot of drafts SS prospects may cost a premium, you can sometimes scoop a CB off waiver wire that will put up comparable or better numbers...

more split alerts : flashing lights & klaxon alarms :

TEN S michael griffin closed the season very strong... in deeper leagues he should be mostly gone, but maybe in smaller leagues, he might have remained unrostered as he began year at CB, & didn't set the league on fire at first... projecting his second half, though, he would have finished significantly higher...

NO CB fred bennett put up huge numbers in second half... don't know a lot about him, or even if he is lock to start (i know they have jason david, who has been pretty disappointing for what he cost the saints)...

CIN CB jonathan joesph strung together a very nice second half... (edit/ADD - with joseph it might be advisable to break down the numbers further & put them under a microscope, & see if the scoring outburst was the result of a few big games with INTs, or more spread out from safer, consistent tackle production... than again, having multiple INT games as rookie could bode well... norton's rookie CB rule applys here, & sometimes stud CBs begun to be passed on less once they become battle tested & proven, which could lead to reduced INTs in future, thus the greater safety & expected reproducibility from steady tackle output)...

ARI CB (& now maybe FS) antrell rolle finished close to top 10 as it was, & won't sneak up on anybody... he had massive, mind boggling second half stats, & would easily have finished #1 among DBs projected over a whole season... his numbers were helped immensely by one game... without looking it up, i think he had 2-3 INTs & 2 TDs (maybe nearly getting a third?), so that would be unlikely to duplicate, but he had quite a few good games in that stretch, so his outburst wasn't solely based on that one game...

* come on, give it up, IDP board

one other thought... it is always good to be aware of guys that were injured, as it can cause them to fall dramatically or all the way off boards of last season stats when trying to screen for these kinds of splits... a good trick is to do deeper screens, time, patience & inclination permitting (i sometimes look 200-300 deep), that will help pick up players that may have not missed whole season, but played part of a season at a high level...

** i find it very helpful to shuttle back & forth between what scouting insights i can glean myself & with help of others (i like to collate scouting reports from places like FBG, coyle, PFW, sporting news, usa today, many other sources), & the actual numbers... i'm not as quantitative likely as some of my counterparts, in terms of projecting numbers for the whole league, & rely more on the scouting side for my projections & rankings... looking at the numbers can temper & augment scouting insights, by grounding them...

the good thing about doing deeper screens, you don't necessarily have to spend a lot of time scouting, & you can outperform opponents & win leagues, not necessarily by being smarter or more insightful, but just by outworking others... & even deeper screens in the off-season done once can only take a few minutes... it can take more regularly in-season, especially if in multiple leagues, but it is easy to do periodically...

the most powerful combo is probably if you couple the two worlds of scouting & tracking numbers (sniffing out splits & other acts of separating information & patterns from noise)... you will be in better position to use background on specific players to INTERPRET the raw numbers, give them context, & render them more meaningful...

for instance, based on the fact that michael griffin & jonathan joseph were highly graded athletes & talents with playmaking potential, & were positioned for success due to 1st round pedigree, shouldn't come as a huge surprise that they flashed promise in rookie seasons... bennett, as i said, i have no idea about... he might be a good candidate to do some cursory research to see if he is likely to be legit or a flash in the pan... rolle is a player who, like griffin & jefferson has 1st round pedigree (actually higher, top 10), but unlike them, had a knee injury & is a three year (?) vet who had generally underwhelmed in his tenure at ARI up to that point... so he should be treated as a separate case from griffin & jefferson... he was however, at one time was VERY highly thought of in the scouting community... IF he is one of the players who doesn't hit the ground running, but at some point later the light does eventually come on, could offer MASSIVE upside... i'm kicking myself for not adding him soon enough in some leagues where available (sure he had been dropped in many leagues, & may still be available in some)... the hazards of being in multiple leagues...

 

[msommer]

I haven't started doing this bit yet but I will have to in the next week or so, getting ready for the UFA auction

 

[CONAN]

Deep league so the pickings are slim but . . .

DE - Jared Devries - DET

5 seasons of pedestrian numbers but came out a bit last year with 27 solos, 3 assisted, 7 sacks, 3 FF, and 3 FR. Had a 3-sack game early against Chicago but otherwise just picked up sacks here and there.



DE - Jarvis Green - DE

24 solos, 15 assists, 7 sacks, 2 FF, 1 FR. He did about the same last year.



DT - Tony Brown - TEN

32 and 20 with 3 sacks and 1 FF. All career highs (ranked 13th in FBGs scoring).



DT - Peko Domata - CIN

36 and 16 with 1.5 sacks and 1 FR (21st).

 

[awesomeness]

Deep league so the pickings are slim but . . .

DE - Jared Devries - DET

5 seasons of pedestrian numbers but came out a bit last year with 27 solos, 3 assisted, 7 sacks, 3 FF, and 3 FR. Had a 3-sack game early against Chicago but otherwise just picked up sacks here and there.



DE - Jarvis Green - DE

24 solos, 15 assists, 7 sacks, 2 FF, 1 FR. He did about the same last year.



DT - Tony Brown - TEN

32 and 20 with 3 sacks and 1 FF. All career highs (ranked 13th in FBGs scoring).



DT - Peko Domata - CIN

36 and 16 with 1.5 sacks and 1 FR (21st).

Nice post, but try not to use such bright fonts, its hard on the eyes.

 

[CONAN]

Deep league so the pickings are slim but . . .

DE - Jared Devries - DET

5 seasons of pedestrian numbers but came out a bit last year with 27 solos, 3 assisted, 7 sacks, 3 FF, and 3 FR. Had a 3-sack game early against Chicago but otherwise just picked up sacks here and there.



DE - Jarvis Green - DE

24 solos, 15 assists, 7 sacks, 2 FF, 1 FR. He did about the same last year.



DT - Tony Brown - TEN

32 and 20 with 3 sacks and 1 FF. All career highs (ranked 13th in FBGs scoring).



DT - Peko Domata - CIN

36 and 16 with 1.5 sacks and 1 FR (21st).

Nice post, but try not to use such bright fonts, its hard on the eyes.

Sorry, after typing it up it all looked the same so I tried to break it up.

 

[obxlegends]

DE - Jared Devries - DET5 seasons of pedestrian numbers but came out a bit last year with 27 solos, 3 assisted, 7 sacks, 3 FF, and 3 FR. Had a 3-sack game early against Chicago but otherwise just picked up sacks here and there.DE - Jarvis Green - DE24 solos, 15 assists, 7 sacks, 2 FF, 1 FR. He did about the same last year.DT - Tony Brown - TEN32 and 20 with 3 sacks and 1 FF. All career highs (ranked 13th in FBGs scoring).DT - Peko Domata - CIN36 and 16 with 1.5 sacks and 1 FR (21st).

Font fix.