What is Pace and Who Dictates It?

The name of the seldom-used blog you are currently reading is called Dictate The Pace. Like most sports-oriented blogs, the name is a feeble attempt at using an otherwise-meaningless, sport-specific phrase to convey to readers the nature of the content. Naturally, I’ve written little on here about dictating the pace, because naming this blog was more of a procedural requirement than a mission statement. But today is different. Today is for pace.

More so than any other major (or semi-major) American sport, pace was long ago fleshed-out in college basketball. KenPom was the first to capture it in any meaningful way nearly 20 years ago, and I (presumably) speak for most CBB fans when I say we probably take for granted how unusual it is for a sport to have such a readily-accessible, catch-all method for projecting the likely game state of any CBB game across the country. Just look at the little game possession projection number on KP’s site and…voila! You now know within a reasonable degree of certainty if the game will be frenetic, average-paced, or a slow grind.

Pace is likewise an extremely important part of any football game, but both the CFB and NFL content industries are eons behind the times in capturing its nuances. Establish the Run’s Pat Thorman writes a weekly column called, appropriately, “Snaps and Pace,” in which he cobbles together component pace metrics (no-huddle rate, pass rate over expectation, plays per minute, etc.) to map out potential play volume in each NFL game. It’s a very simple concept, and yet, it is among the most widely-read columns in fantasy/betting circles because (as far as I know) he was the first among NFL or CFB writers to distill this data into an easily-digestible piece of weekly content.

Betting Totals
Earlier in the offseason, I started digging into whether there was something more that could be done with pace in CBB. After all, little has changed in the 20 or so years in which pace data has been available on KenPom and Torvik. Are we really at end of the discovery road with the entire concept of pace? The nerds in baseball analytics invent some ridiculously cool new metric seemingly every other week (Bat path angle? Swing path tilt? This shit is cool as hell). Where are “our” fancy new discoveries? Is it just plain old Four Factors ‘till we die?

At some point (I don’t recall when), KenPom began to delineate between offensive possession length (APL) and defensive possession length, which was (and still is) a very helpful dichotomy. But, for a betting person, this was just another ingredient in a caked that was already baked. Oddsmakers long ago began co-opting KP’s numbers (particularly on mid-major totals), which meant that there was no edge to be gained from strictly adhering to KP’s methodology.

The real reason I’ve invested countless hours this offseason into this bottomless pit of despair is that my totals have been a source of incessant aggravation over the past two seasons, lagging woefully behind my spreads (or sides, for the well-versed) over a very large sample (data from BetStamp spanning the 2023-24 and 2024-25 CBB seasons):

Despite nearly identical CLV (useful signal in a huge-sample data set like this), my ROI on sides is nearly 3.5x my ROI on totals. The obvious answer is, “hey JAF, maybe you’re just not that good at totals?” And sure, maybe it’s that, but I’m a headstrong motherfucker, and sometimes, I need to touch the hot stove more than once or twice to make sure that it’s really the reason my hand is now burning.

After spending the first few months of the offseason messing around with rosters and trudging through the thicket of player fit and construction, I’ve taken a step back recently to build something new for attacking the totals market. And I’m using the term “build” quite loosely here. I was a terrible math student who nearly failed out of high school geometry and can’t write a line of basic code. There’s a reason I’m a lawyer.

A New Approach
The Tempo Imposition System (TIS) seeks to answer an elementary question: Which team will actually control the way a game is played? Every team has an identity (pace-and-space, halfcourt grind, token pressure falling back into zone, etc.), but tempo on an individual game basis is decided by whichever team can impose its style once the game starts. Announcers will often make obligatory reference to “controlling the tempo,” but the hell does that mean?

The heuristic tries (perhaps futilely; we’ll find out soon enough) to measure that control empirically. It breaks tempo down into two complementary scores: a Fast-Impose Score (FIS), which measures how well a team speeds games up, and a Slow-Impose Score (SIS), which measures how well a team slows them down. Each team (no matter how fast or how slow) receives an FIS and an SIS. Together, those two tell us whether a team is a true pace dictator, a flexible chameleon, or is at the tender mercy of its opponent’s preferred play style.

To do that, the TIS uses basic, widely-available stats that describe behavior and control. Instead of counting possessions, it looks at how possessions unfold: how early teams shoot (average possession length), how many shots they create per trip, whether they extend possessions with offensive rebounds, and whether they give them back with turnovers, parsed between steals (live-ball) and non-steal TOs (dead-ball).

On defense, it tracks the ability to force long possessions, end trips with clean rebounds, avoid fouls that stop the clock, and prevent opponents from getting clean looks. Each of these pieces represents a small share of “control” over how fast or slow a game feels. Add them up—and adjust for schedule and opponent—and you get a single z-score showing how much a team can bend pace to its will.

What the heuristic is not doing is predicting the exact number of possessions or points. It’s not an offensive efficiency model. There are no eFG% or shooting-related inputs in here. Whether the offense can score or whether the defense can limit scores is irrelevant in this project. Instead, it focuses purely on game state — the flow of time, volume of shots, and turnover texture. All shot-making noise is stripped out.

My hope is that this approach is able to catch early-season signals faster than my “traditional” methods of handicapping totals. Raw possessions per game are descriptive—they tell us what’s happened—but they don’t say who drove the pace of play. Two teams can both average 70 possessions, yet one did it by forcing its opponents to run, while the other simply mirrored whomever they played. This approach separates those concepts. When a high-impose team meets a passive mirror, we can anticipate that the pace will be driven by the high-impose team (FAST or SLOW) with the passive (low-imposing) team just going along for the ride.

The Dictators Dictaters
To demonstrate, the two screenshots below show the top 10 highest-rated (z-scored) dictaters (pace imposers) and the 10 lowest-rated (z-scored) pace followers. The Dictater/Imposer group is, unsurprisingly, mostly strong teams (Nevada was an analytical quandary last year that I still can’t explain). The “bad” group is, generally, teams who sucked.

Just out of screen in the first screenshot are #11 Alabama, #14 Houston, #15 Florida, #24 Auburn, #28 Iowa State, #29 Arizona, #30 Texas Tech, and #38 Maryland.

The (massive) outlier among top four seeds was Purdue, sandwiched in between #241 Georgia Tech and #243 Fresno. I’m not sounding alarm bells or anything, but it’s notable that this was the one nationally elite team (using top 4 Tourney seed as a proxy for “elite”) that was below-average in imposing its “will” on the game (Wisconsin was second-worst at a distant #167) and is now running back a similar roster for 2025-26.

Was St. John’s the “best” FAST team in the country last year? Probably not. But they were the team likeliest to impose their brand on any individual game.

Likewise, Clemson wasn’t the “best” (from a powers rating standpoint) SLOW team nationally, but they forced nearly everyone to play Brownell’s style.

The Top 10 Dictaters: In One Sentence

• St. John’s – Relentless tempo accelerator with ultra-short offensive APL and live-ball turnover creation; forces opponents into faster possessions than they want.

• Clemson – Textbook slow-control team: long offensive APL, half-court patience, and defensive schemes that cut transition frequency to a crawl.

• UCLA – Possession suppressor; elite defensive discipline and length extended opponent possessions, producing consistently low-tempo, low-event games.

• Nevada – Another slow dictator built on ball-control guards and a no-gamble defense; very few steals either way, but every trip was a chore.

• Marquette – Organized chaos: high-pressure perimeter defense, quick early-clock offense, and frequent turnover-to-transition conversions drove elevated possession counts.

• UAB – Physical, rim-pressuring group that thrives in up-tempo exchanges; big offensive-rebound rates and pace off misses yield short possessions.

• Gonzaga – Trademark fast-break orientation; strong rebounding to trigger transition plus early-offense spacing that shortened APL on both ends.

• Utah State – Deliberate pace governor; half-court efficiency, patient ball movement, and defensive structure minimizing quick shots via shifting defenses make them a pace controller.

• Duke – Half-court technician that dictates pace through control of shot selection and rim defense, limiting run-outs and keeping tempo steady at its preferred rhythm.

• New Mexico – High-possession attacker led by guard speed and turnover pressure; thrives in up-and-down games and drags slower teams into faster rhythms.

You may have noticed a column on the far-right hand side in the above chart labeled “Net TIS (Higher = Overs).” This is a basic FIS (Fast Impose) minus SIS (Slow Impose) calculation. Remember, there are no eFG% or shooting calculations in here.

Importantly, a team with a strong SLOW imposition rating isn’t necessarily a good defense. Northern Kentucky finished with the 25th-best imposition rating (SLOW), but because they couldn’t keep the ball out of the net (#252 in opponent eFG%), their raw game totals trended higher than would be expected from their game environments. Virginia was another (and maybe the best) example of this “phenomenon”: the game environments were brutally slow, but the Cavs couldn’t string together stops (#172 opponent eFG%) and they became an “over” team because of it.

As a spot check to make sure this was working as intended, I calculated as a test the 10 highest and 10 lowest net TIS ratings from last season. I was prepared to scrap this entire project if Drake wasn’t the lowest (slowest) net TIS:

Fastest? Alabama, UAB, Bryant. Check.
Slowest? Drake, Wagner, Virginia. Check.

Adding in Shot Quality
I thought about leaving this as it was, and for awhile, that’s what I did, but I was intrigued by the possibility of layering in *something* else to capture shot selection and/or shot diet without injecting outcome-based shooting stats. Otherwise, wasn’t I just reshuffling the KenPom/Four Factors deck and slapping a fresh coat of paint on it? More or less, yes. I settled on using ShotQuality’s shot selection (offense + defense) and rim & 3 rates (offense + defense) grades to create an additional “blended” metric that pairs with the FIS and SIS ratings and gets me away from a purely Four Factors-based build (like anything else, SQ has its warts, but after years of messing with the data, I’m convinced that it has quite a bit of standalone value and is among the few analytics sites not rooted in Four Factors methodology).

To avoid the SQ data swamping the FIS/SIS ratings, I weighted FIS/SIS 65% and SQ 35%. To quantify everything in one catch-all number, I created an “Environment Score” with this blended 65/35 weighting:

Winthrop had the best SQ Shot Strength score for overs (elite offensive shot selection and rim + 3, and near-equally bad rates of same allowed defensively), but the Eagles were “only” eighth in overall pace environment because their FIS/SIS profile was merely strong but not overwhelming. Instead, by a considerable margin, Alabama rated as the most scoring-prolific overall (blended) pace environment nationally. I know, you’re floored by this. Alabama OVERS! Who knew? Intel like this is why you subscribe to this free blog.

UAB is a fascinating case. The Blazers were tied for second (behind Alabama) in net TIS but finished 132nd in net SQ shot strength (a proxy for net shot quality taken/allowed). They also often just could not shoot — a partial byproduct of taking a lot of difficult mid-range shots. This is the most important distinction with this particular metric: it is showing us the best “environments” in which pace will thrive (or be throttled), but it can’t account for game-to-game fluctuations in eFG% or the “true” ability of the offense or defense to put the ball in (or keep it out of) the basket. Because UAB finished 6th overall in Imposition Rate (the “dictating” metric), they were able to speed up a lot of slower or passive mirror teams.

Houston had the worst Net SQ Shot Strength score for overs. If you’ve watched Houston games (which all of you have, obviously), you know what this looks like. We’re what, eight (maybe nine) years into the Kelvin Sampson thing? The Cougs dick around with the ball for awhile in the halfcourt, wait for a mismatch, take terrible shots (359th in offensive shot quality), make a higher-than-they-should percentage of those terrible shots, and clean up a shitload of their own misses by back-tapping the ball for catch-and-shoots. At the other end, their opponents seldom get a good look against Houston’s obnoxious defensive closeouts and pterodactyl length. Thus, it’s not hard to see why they graded out as the most unfavorable SQ team nationally for points scored and as the worst scoring “environment” overall among Big 12 teams when blending FIS/SIS and the SQ data.

Tarleton is the closest thing to a “model-breaker” as there is in CBB. A reasonable person could assume that a team that plays like Tarleton (in-your-shorts pressure, all-out denial defense) with obnoxious, unwatchable two-way fouling would “dictate” the terms of the game…but they don’t. Yes, they play (very) slow, but they don’t force the opponent to play that way. In fact, they allowed the FASTEST defensive APL in the country. Totally bizarre team and style (and has been throughout Billy Clyde’s tenure).

---

Thanks for reading. I assume none of it made any goddamn sense.

Three weeks.