He included fumbles that players made while on special teams' duty.
Special teams' fumbles were with "neutral" footballs NOT supplied (read "doctored" or potentially so) by individual teams. Special teams involve footballs from a distinct set-aside pooling of such balls.
Therefore, I "fixed" that by conducting my own analysis -- removing special teams' fumbles.
Mine was also a "tad" more comprehensive (than Sharp's) because my analysis includes players like Kevin Faulk, who had over 430 "touches" post 2006...and hence, makes for a worthwhile comparison to how he did after 2006 vs. 1999-2006 seasons.
So, here was MOST of my Feb. 25 Freeper Vanity I posted:
*****************************************************
How might a statistical analyst give his best case that leprechauns are indeed at large in New England?
Well, what if I told you that the Patriots' defense fumbled the ball...
...more often in (take your pick which of these following seasons) 2003, 2002, 2001, 1999 than the Patriots' running backs did rushing the ball in 2007 even including all three playoff games?
...Or their defense fumbled the ball as often in 2001 as the Patriots' running backs did rushing the ball for the entire 2007 and 2008 seasons combined (35 games including three playoff games)?!
...Or their defense fumbled the ball as much combining regular seasons 1999-2003 as did the Patriots running backs' rushing the ball combining regular seasons 2007-2008 and 2010-2011?
(Now you know what the job description of a New England leprechaun is every pre-game!)
Beyond that, just compare the 2006-2007 seasons and broader patterns (either 2000-2014, or 2003-2014)
Category | 2006 NE Season | 2007 NE Season |
Overall fumbles | 31 (27 regular season) | 17 (14 regular season) |
Rushing fumbles by Patriot running backs | 7 (19 games, including 3 playoffs) | 0 (19 games, including 3 playoffs) |
Fumble rate per game | Avg team: 1.5 vs. NE's 1.6 | Avg team: 1.6 vs. NE's 0.8 (Less fumbles by half!) Note: Even indoor based teams averaged 1.55 fumbles per game) |
Brady's Completion % | 61.8% | 68.9% |
Fumbles by Teams Per Game | ||||
2003-2006 NE | 2003-2006 Other 31 teams | 2007-2014 NE | 2007-2014 Other 23 Outdoor teams | 2007-2014 Indoor-based teams (8) |
1.46 | 1.6 | 0.96 [this is improvement of 1 less fumble every 2 games vs. previous NE teams] | 1.46 [this means one more fumble every 2 games than NE] | 1.29 [this means one more fumble every 3 games than NE] |
So, for certain NFL attorneys who may want to delve into local Boston lore, what pot o' gold nuggets of evidence seems to suggest shenanigans on the loose going back about 8 years?
Note: before assessing chart below, it might be of help to review Warren Sharp's original chart on 19 of these players below:
* Jan. 28, 2015 update: New England Patriots Fumble More Often When Playing for Other Teams)
* See also: January 22, 2015: The New England Patriots Prevention of Fumbles is Nearly Impossible and...
* January 23 2015 Slate: Dumb Luck: The New England Patriots prevention of fumbles is nearly impossible.
Indicators of 'Deflategate' & 'Ballghazi' as Urban Legend |
Red Flags in Pat Stats Suggesting Shenanigans |
1. Warren Sharp's analytics case of embellished fumbles (what was he thinking or not thinking anyway?) Sharp treated all fumbles as equal & relevant research. But, alas, they aren't. Simply put, if a team is accused of doctoring their own footballs, & if special teams use a common pool of balls providing no competitive advantage, then special teams' fumbles are irrelevant & need special segmenting from all analytical charts. This impacted Sharp's charts how? One Sharp chart lists 19 players who were either former Patriots or played elsewhere prior to coming to New England. These 19 lost 124 overall fumbles. The problem is three dozen fumbles occurred during returning a punt or kick. (That's 29% of fumbles in list). | 1. The however to this is it's an equal-opportunity application mistake: On the other side of the ledger those tracking Pat fumbles 2007-2014 9 of 39 fumbles were likewise special teams (23% vs. 29% on other side). In other words: Most of fumbles removed from the balance sheet prove to be a wash. All it does is to heighten the number of touches on each comparison side per fumble. It is true -- for sake of only including 'relevant' fumbles as it applies to this case study -- that Sharp's "44 touches per fumble" & "73 touches per fumble is a myth. The actual touches are much higher on both comparative sides. Also, when playoff stats are added to Sharp's charts along with one additional measurement RB Kevin Faulk it fleshes out an even a greater measurement: The lopsided touches per fumble ratio Sharp arrived at 98 67 among those 19 players & 107-53 among the five players with 300+ Patriot touches...grows to 145-87 among 20 players & an astounding 190-70 split among 8 players with 274+ touches (Wes Welker, Laurence Maroney, BenJarvus Green-Ellis, Faulk, Danny Woodhead, Sammy Morris, LeGarrette Blount, & Randy Moss) |
1a. How did (1) above play out? Wes Welker, for example, returned punts/kickoffs for both Pats & other teams: Therefore, half of his Pat fumbles were irrelevant to case study & ALL of his non-Patriot fumbles were likewise irrelevant. Same with Brandon Tate re: his non-Patriot fumbles (all irrelevant). One would think that including Tate's 11 fumbles in only 35 touches would greatly skew the results to work against Sharp's hypothesis. A dozen other fumbles were likewise removed on the non Pats' side: (Amendola, 5; Moss, 3; Stallworth, 2; + Jordan & Gaffney, 1 apiece). | 1a. How the above played out on Pat stat side '07-14: Welker returned punts & kicks for the Pats, too: So half of his fumbles were special teams' related. Amendola & Morris also had each had a special teams' fumbles removed. And tho Sharp didn't include Kevin Faulk in his chart because Faulk only played for the Pats, Faulk is perhaps THE most interesting case study, but not for special teams' sake (Faulk had one special teams' fumble removed from his stat total; beyond that, he only fumbled once in his last five seasons with the Pats - & it was a reception, not a rush. By comparison, in the alleged pre-Ballghazi era, Faulk fumbled it 24 times (4 special teams) over eight seasons: 13 rushing, 7 after catches. IoW, he averaged 1 'relevant' fumble every 35 touches thru 2006; suddenly it mushroomed to 1 'relevant' fumble every 433 touches 2007-2011. IoW, Kevin Faulk himself is the face -- the poster boy -- for 'ballghazi shenanigans'! |
2. 'Relevant' fumbles & fumble ratios: When the raw fumbles #s are scrubbed & only 'relevant ones remain, 8 of 19 players Sharp analyzed don't match the we fumbled more wearing non-Patriot shirts narrative: Danny Woodhead, Fred Taylor, Brandon Lloyd, Brandon Lafell, Deion Branch & Lamont Jordan all have similar fumble ratio numbers no matter which team they've played for; + Wes Welker, Randy Moss when properly stripped of those special teams' fumbles even showed significantly more of a penchant to fumble when playing for the Patriots. | 2. Collective stats for 11 Patriots 2007-2014 show only 1 fumble every 472 touches! While some of the Patriot fumble miserliness 07-14 are indeed attributable to guys who tend not to fumble often (beyond special teams at least) Welker, Woodhead, & Laurence Maroney. Yet when the other 15 Rbs & Wide-outs are surveyed, 'twas an almost impossible scenario to look @ the stats of 11 of them & realize these 11 combined for almost 1900 touches between them during those 8 seasons, & yet they fumbled only four times: 1889 touches & only four collective fumbles by Green-Ellis, Faulk, Branch, Taylor, Amendola, Lloyd, Lafell, Jordan, Gaffney, Evans, Stallworth that's only one fumble per 472 touches |
3. A fumble-by-fumble review turns up that the players most responsible for Pat fumbles were quarterbacks! NFL Fumble Pie is cut up into 5 pieces: Fumbles by Qbs, Rbs, Receivers, Special teams, & the occasional post-interception fumble. For 1999-2006, Pats Qbs made 45% of fumbles; that was reduced to 33% 2007-2014. Sharp (& others including myself) don't want to include Qbs for analysis purposes because they already tend to have an untucked ball in most play situations. The key point here, though, is between 07-14, Pats Qbs + special teams accounted for over half of all team fumbles, leaving less room to shenanigize anything | 3. The so-called flip side of this argument is actually the same argument: Yes, review the fumbles to see who was actually making them, or rather, no longer coming even close to making them. How is it that the Patriot Rbs averaged less than 3 fumbles per season 2007-2014? How is it than when you include playoff games, the Pats average a rushing fumble by a RB or wide-out about once every six games? How did the Pats go through '07 19 games including playoffs without its Rbs managing to fumble the ball on a rushing play? (Kevin Faulk had one fumble but even that came on pass he caught) |
4. If you're trying to explain why Brady had his sudden 2007 completion % surge, look no further than Randy Moss. Moss had previously had two 100+ reception years; his new presence in '07 accounted for 98 receptions. | 4. Indeed, personnel are very important considerations. And it's personnel breakdown charts like the one below a corrective revision of Warren Sharp's listing 19 players that makes the 'Ballghazi' case. (Note: added Kevin Faulk as a 20th player because Faulk had over 430 touches post 2006 & that can be readily compared to his 1999-2006 Patriot seasons where he had 700 touches) that also provides a face to this ghost of Ballghazi. |
Fumbles-per-touch analytics | ||
STATISTICAL BOUNDARIES | COLOFORNIAN'S ANALYTICAL BOUNDARIES | WARREN SHARP'S ANALYTICAL BOUNDARIES |
Are only 'relevant' fumbles -- non-special teams' fumbles -- included in fumbles per touch comparisons? | YES | NO (Sharp's are accumulative) |
Are playoff stats included in 'relevant' fumbles by touch comparisons? | YES | NO (Sharp uses only season stats despite very controversy arising in playoff context) |
Are pre-2007 Patriots stats included in 'relevant fumbles by touch comparisons? | YES | NO (Sharp didn't include Kevin Faulk as he only played for NE; & he didn't use Deion Branch, Ben Watson & Laurence Maroney pre-2007 Pat stats: Yet these stats are relevant for comparison sake) |
Left side: NE PLAYER: 07-14 | Right side: NON-NE PLAYER + NE PLAYERS PRE-2007 | |||||||||
NAME | RECEPTIONS | RUSHES | TOUCHES | 'RELEVANT' FUMBLES* | 'RELEVANT' FUMBLES PER TOUCH | RECEPTIONS | RUSHES | TOUCHES | 'RELEVANT' FUMBLES | 'RELEVANT' FUMBLES PER TOUCH |
Wes Welker | 741 | 21 | 762 | 6 relevant (other 6 on special teams) | 1 per 127 | 237 | 1 | 238 | 0 relevant (all 13 on special teams) | Less than 1 per 238 |
Laurence Maroney | 45 | 644 | 689 | 5 | 1 per 138 | 5 | 67 | 72 | 3 | 1 per 24 |
BenJarvus Green-Ellis | 31 | 557 | 588 | 0 | Less than 1 per 588 | 28 | 517 | 545 | 5 | 1 per 109 |
Kevin Faulk | 181 | 252 | 433 | 1 relevant (1 special team) | 1 per 433 | 301 | 700 | 1001 | 20 relevant (4 special teams) | 1 per 50 |
Danny Woodhead | 104 | 285 | 389 | 3 | 1 per 130 | 160 | 95 | 255 | 2 | 1 per 128 |
Sammy Morris | 52 | 335 | 387 | 3 relevant (1 special team) | 1 per 129 | 117 | 402 | 519 | 8 | 1 per 65 |
LaGarrette Blount | 6 | 275 | 281 | 3 | 1 per 94 | 27 | 491 | 518 | All 10 relevant | 1 per 52 |
Randy Moss | 271 | 3 | 274 | All 5 relevant | 1 per 55 | 765 | 23 | 788 | 8 relevant by comparison (3 special teams) | 1 per 98 |
Deion Branch | 130 | 0 | 63 | 0 | Less than 1 per 63 | 452 | 11 | 463 | 1 relevant (2 special teams by comparison) | 1 per 463 |
Fred Taylor | 4 | 108 | 112 | 1 | 1 per 112 | 293 | 2555 | 2848 | All 26 relevant by comparison | 1 per 110 |
Danny Amendola | 90 | 2 | 92 | 0 relevant (1 special team) | Less than 1 per 92 | 196 | 12 | 208 | 5 relevant by comparison (5 special teams) | 1 per 42 |
Ben Watson | 91 | 1 | 92 | 2 | 1 per 46 | 210 | 1 | 211 | 6 | 1 per 35 |
Brandon Lloyd | 86 | 0 | 86 | 0 | Less than 1 per 86 | 325 | 1 | 326 | 4 | 1 per 82 |
Brandon Lafell | 83 | 2 | 85 | 1 | 1 per 85 | 171 | 7 | 178 | 2 | 1 per 89 |
Lamont Jordan | 0 | 80 | 80 | 1 | 1 per 80 | 163 | 856 | 1019 | 7 relevant (1 special team) | 1 per 146 |
Jabar Gaffney | 78 | 0 | 78 | 0 | Less than 1 per 78 | 394 | 9 | 403 | 3 relevant (1 special team) | 1 per 134 |
Heath Evans | 8 | 48 | 56 | 0 | less than 1 per 56 | 59 | 121 | 180 | 3 | 1 per 60 |
Donte Stallworth | 55 | 1 | 56 | 0 | Less than 1 per 56 | 280 | 20 | 300 | 3 relevant (2 special teams) | 1 per 100 |
Brandon Tate | 24 | 6 | 30 | 1 | 1 per 30 | 31 | 4 | 35 | 0 relevant (11 special teams by comparison) | Less than 1 per 35 |
TOTALS | 2080 | 2620 | 4700 | 32 | 1 in 145 | 4214 | 5993 | 10107 | 116 | 1 in 87 |
8 Players' collective totals who had 274+ touches for Patriots | 1431 | 2372 | 3803 | 20 | 1 per 190 | Non-NE or Early NE | 1640 | 2296 | 56 | 1 per 70 |