2019 Position Group Progress Report: Safeties

Courtesy iuhoosiers.com

 

Over the next month or so, we’ll analyze the performance of each IUFB position group during the 2019 season. Today, we start with IU’s safeties, probably the position group that struggled the most this season.

Expectations

Going into 2019, expectations were relatively high for IU’s safeties. Four-year starter Jonathan Crawford’s graduation left a hole at field safety, but four safeties returned who saw significant action in 2018: Bryant Fitzgerald, Devon Matthews, Juwan Burgess, and Khalil Bryant. It was assumed that Fitzgerald would start at boundary safety1, while Matthews would start opposite him at field safety2 Burgess and Bryant would still see plenty of time at FS and BS, respectively. Bryant was a senior, while the other three were all sophomores. Fitzgerald in particular was expected to be a playmaker. He flashed his potential while starting three games at BS3 in his redshirt freshman season.

Performance

Then the season happened. Right off the bat, there were issues. In the season opener against Ball State, Fitzgerald gambled on a rotation following a corner blitz, dove for an INT or at least a PBU, and…missed. Thirty-six yards later, Ball State had cut the IU lead to 7 midway through the 4th quarter. In the next game, Devon Matthews was beaten badly on an OSU TD. Against Michigan State, Fitzgerald dropped a likely pick-six on Sparty’s first possession and later gave up two TDs through the air. By late October, Khalil Bryant had supplanted Fitzgerald as the starting BS, while Burgess took over for Matthews at FS. This was always going to be a platoon at both safety spots, but instead of being 60-40 or 70-30 in favor of the initial starters (Fitzgerald and Matthews), it became a similar split but with the Bryant and Burgess leading the way. Here’s the season-long playing time breakdown:

A couple notes about the chart above: (1) the “poss” field is the total number of possessions that each player appeared in. A defender need only play a single play during an opponent’s offensive series to count; and (2) the “%” field shows the percent of the total non-garbage time possessions in which the player participated. Here, the possession breakdown is a little misleading. Throughout the season, on passing downs (generally third and 5 or more), IU went to a specific personnel grouping that involved Devon Matthews playing boundary safety and veteran CB Andre Brown lining up at field safety.4 That means Matthews was given credit for appearing in many possessions where he only played a play or two on passing downs, particularly later in the season. If we had a count of total snaps played, I expect Burgess and Bryant would have more snaps played than Matthews.5

But swapping in Bryant and Burgess didn’t fix the underlying problem. However Wommack & Co. divvied up playing time, this was a group of safeties who struggled to limit big plays in the passing game while failing to make any big plays of their own. First, the chart below shows how this group fared against the pass:

Woof. Matthews’ coverage numbers are respectable, but it’s not pretty for the other three. For Fitzgerald, allowing a TD on a quarter of targets faced is less than ideal. Beyond that, it’s not a good sign when the duo that’s considered the better option (Bryant and Burgess)¬†allows a combined 262 yards on 19 targets (13.7 YPT). In the Maryland, Nebraska, Penn State, and Michigan games, one or both of Bryant and Burgess was victimized by a big passing play. And the whole crew broke up just 2 passes and failed to record a single interception.

Which leads us to the second and probably more important point – in 2019, IU simply did not get any playmaking out of the safety position. Bill Connelly popularized a defensive stat called “Havoc Rate.” The premise is to count up stats that indicate a defender causing havoc – tackles for a loss, forced fumbles, passes defensed – and divide that by the total number of plays to arrive at a measure of how aggressive or playmaking a defense or given defender is. Connelly also mentions that “if QB hurries were a reliable stat (at the college level, there is far too much inconsistency in how they are recorded), they would be included here, too.”¬† While QB hurries are objective based on different official scorers across the country and thus unreliable in comparing teams nationally, I’ve added it because I track QB pressures for Indiana so there should be a good level of consistency in my tracking. As a result, the chart below shows usual the components of Havoc, but “TFL” does not include sacks, which are captured in pressures, shown in the “Press.” field. This chart shows how IU’s safeties have fared at creating havoc since Tom Allen took over as defensive coordinator in 2016.6

There’s a quite a bit to unpack in this chart, but the gist is this: IU safeties were reasonably consistent at creating havoc from 2016-2018, largely due to mainstay Jonathan Crawford. That fell off a cliff in 2019. Devon Matthews somehow managed to not record a single havoc stat, despite appearing in 77 possessions. Bryant Fitzgerald tallied 8 havoc plays in 2018 – a respectable total for a redshirt freshman – only to drop off to 5 in 2019. If nothing else, we can credit Khalil Bryant for his consistency – he followed up 2018’s 2 havoc play performance with 1 havoc play in 2019, despite a significant uptick in playing time.

Part of this is likely a result of scheme. For example, by my count, IU’s safeties were called upon to blitz only 10 times this season. It’s pretty hard to rack up QB pressures at safety when you aren’t blitzing. Moreover, it’s possible coaching is encouraging these guys to play more conservatively, which limits opportunities for breaking up passes and making tackles in the backfield.

But it ain’t all coaching. Jonathan Crawford played for the same coaches and he never had fewer than 8.5 havoc plays in a season, and in 2016 and 2017, he personally created more havoc than the 2019 quartet combined. It’s no coincidence that Crawford was the best athlete to play safety at IU in the 16 seasons I’ve been paying attention. General athleticism certainly seems to lead to playmaking, and there doesn’t appear to be a Crawford-level athlete in the current group. Beyond athleticism, I’d suggest that there’s also a confidence component. It stands to reason that knowing what you’re doing, expecting to make plays, and generally playing with confidence can lead a defender to create some havoc. Just watch Tiawan Mullen play corner. On this point, from my (admittedly distant) vantage point, it seemed pretty obvious that Bryant Fitzgerald’s early season struggles zapped his confidence. By the end of the season, he was a shell of the instinctive, powerful, playmaking safety we saw glimpses of in 2018.

Just so you don’t think it’s all bad news, this group actually tackled fairly well:

For a position where most tackle attempts are in the open field, any tackle percentage above 90% is quite good. For reference, Jonathan Crawford only topped 90% once in the three seasons I tracked his tackling. In the Nebraska game in particular, Khalil Bryant was everywhere, cleaning up his fellow defender’s mistakes and stopping 8 or 10 yard plays from becoming 20 or 40 yard chunk plays. Not all safeties need to be hard-hitting thumpers to be effective, and at his best, Bryant put himself in the right place and made the tackle, particularly against the run.

The Future

Of the 2019 quartet, only Khalil Bryant is graduating. The other three will all be juniors in 2020, which means they still have plenty of time to improve. In terms of who might join them and/or cut into their playing time, freshman Josh Sanguinetti is probably the leading candidate. He redshirted in 2019 (and I believe he suffered an injury in practice midway through the season), but he reportedly turned some heads in Fall Camp and during early season practice. Noah Pierre, who will be a redshirt sophomore in 2020, appeared to be the 5th safety in 2019 and could challenge for more PT. It’s also entirely possible that recent commit Bryson Bonds could work his way into the rotation as a true freshman.

On top of those options, Kane Wommack and his staff probably shouldn’t rule out a position change as they try to find an answer at safety. Cam Jones worked at safety during Spring Practice in 2019, but I think he’s probably a little too big to be effective, and he’s likely to start at WLB in 2020. While Reese Taylor seems to have the physicality to handle safety, he is probably more valuable at CB. If I had to come up with a position change to help the safety position, I would recommend current second-team Husky Jamar Johnson. It’s unlikely he’ll bypass Marcelino Ball to start at Husky in 2020 (Ball’s final year of eligibility) – even though he has absolutely narrowed that gap. In brief flashes, he’s shown ball skills, and the athleticism is there. Also, former four-star recruit Cam Williams, who redshirted in 2019, should be ready to back up Ball at Husky if Johnson moved to safety. Not saying it will happen – just noting that this may be a way to get IU’s best defensive players on the field more.

All that said, the best thing for the safety position is for the trio of sophomores, most notably Bryant Fitzgerald, to get back on track and begin to realize their potential. Adding a confident, playmaking Fitzgerald to the rest of the returnees from the 2019 defense would be a significant boost.

  1. The safety aligned to the boundary, or short, side of the field.
  2. The safety aligned to the field, or wide, side of the field.
  3. And one a Husky, due to Marcelino Ball’s targeting suspension.
  4. I didn’t break out Andre Brown’s appearances by position so the CB possession breakdown may be a little misleading as well. Look folks, you get what you pay for here.
  5. I’ll get to that when this site starts making enough for me to quit my day job…
  6. This chart also has a little garbage time/non-garbage time conflict because the official stats (TFL, FF, PD) are all for the entire season, which includes the Eastern Illinois game and garbage time, while the pressure stat is strictly non-garbage time. While the impact should be pretty minimal, there is a difference, as shown in the 4 passes defensed total for 2019, as opposed to the 2 PD in the chart above which was non-garbage time only.