An Explanation of the new FLW (Feud Length in Weeks) Statistic

In January 2022 we launched a new statistic called FLW, which stands for Feud Length in Weeks. It is an attempt to “measure” feuds in professional wrestling during the territorial era in a way that the output is easily understandable. This post goes into detail about how this stat was created, why it was created, what it can and can’t do, and how to best interpret and analyze it.
What is FLW?

Feud Length in Weeks is a statistic that approximates the length of a feud, based on the overarching principle that if a feud was successful (i.e., drew money), promoters and bookers would continue it for as long as they could.

It is expressed in “weeks”. This is not an actual measure of how long the feud lasted, but merely an approximation based in aggregate on how many matches the two had against one another in a certain period of time as a percentage of the total number of shows the wrestlers worked. It also factors in where on the card the match takes place, so a matchup that usually happens in the main event will have a higher FLW than one that happens second from the top.

As a very simple example, let’s say a territory runs one show per night, using all the wrestlers in the crew, six nights a week. If the main event is the same every night, then the FLW for that match would be 1.00. If a particular main event happened three nights one week and three nights the following week, it also would have an FLW of 1.00. It doesn’t matter if the matches in the second week were rematches in the same towns as the first week or if they were ‘first-time’ matches in the other towns, the FLW would be the same. The idea is if the match draws, they’ll run it as frequently as they can wherever they can.

A great example of this is the Jerry Lawler vs Bill Dundee feud from 1977. When the two were first put together, Jerry Jarrett probably had no idea how long the feud would last (11 straight weeks in Memphis). But the feud kept drawing fans each week far above what they normally were drawing (at that time in that town), so Jerry would have been a fool to blow it off prematurely. Thus, he kept booking rematches with escalating stipulations (the promotion’s Southern Heavyweight title was usually at stake, along with a Cadillac that was Dundee’s in real life). And as the feud kept drawing well past a time when other feuds usually dried up, he had to get more creative with the stipulations, with one wrestler’s hair being at stake while the other wrestler would put up the title and/or the Cadillac. Eventually, Lawler’s manager, Mickey Poole, had his hair at stake, and when Lawler lost that match, Poole got his hair cut. But the feud was still drawing well. In fact, even after what was supposed to have been the “final blowoff” occurred, with Lawler announcing ahead of time he was going to retire after the match and both Lawler and Dundee putting their hair on the line (and in a swerve since Lawler had announced he was retiring, it was Dundee who actually lost the match), Jerry realized the feud still had some life left. So he arranged for Dundee’s wife to put her hair up, and for Lawler to delay his retirement (which was actually part of an angle to turn him babyface) one more week. The last match in the series saw Dundee’s wife lose her hair in exchange for a $3,000 payoff. With ticket prices averaging between $3 and $4, this means Jarrett was confident that the match would draw significantly more fans than if he had a different main event. The attendance for that card, on September 13th, was just over 8,000 and the gate was over $28,000. The following week, without Lawler vs Dundee on top, they drew less than half that (3,167 paying $7,998 with reduced ticket prices). So the $3,000 paid to Beverly was a wise investment, as it led to around $20,000 in ticket sales above and beyond the baseline.

While the feud ran for 11 straight weeks in Memphis, it did not run as often in the other towns. In both Louisville and Evansville, it went for eight weeks. In the smaller towns on the circuit (including Blytheville, Jonesboro, and Jackson), it ran far less often. Around this time, Lawler did not work a full schedule, hitting the big towns regularly and making infrequent appearances in the smaller towns. So when you take all that into consideration, the FLW for their feud is going to be lower than 11, and likely lower than 8, since it wasn’t happening in the small anywhere near as often. In fact, the FLW for the Lawler vs Dundee feud in the third quarter of 1977 was a 6.72. While that might seem low, that is one of the larger FLW scores I’ve seen in a three-month period of time. Throughout the month of January, I’ll be listing some of the higher FLW scores I’ve encountered so far in my research. Keep in mind that I haven’t researched all territories for all time periods. But I have done enough that I feel confident in the results.

Now that I’ve discussed what FLW is, let’s delve into how I came up with it.

My various attempts at quantifying feuds

The FLW score is my third attempt to “measure” feuds. The first was a statistic I called a FEUD (Frequent Encounters Using Data) Score. It was displayed as a whole integer. If the SPOT Rating statistic I have been using was the equivalent of Batting Average in baseball, the idea was for the FEUD Score to be the equivalent of Home Runs. The scaling was similar to home runs as well, with 25 or over being pretty good and 40 or above being extremely good.

The FEUD Score didn’t consider *where* a match happened on the cards, so every now and then a mid-card matchup would make its way into the higher FEUD Scores. So I began the process of developing a new statistic that would take this into account. Additionally, this new statistic would account not just for singles matches between two wrestlers, but also times they were on opposite sides of a tag team match, or a six-man. The new statistic, which was called Frequent Opponents, was displayed as a percentage. The idea was to show how often Wrestler A was on the opposite side of a match, any type of match, with Wrestler B. I also showed these percentages week-by-week, with the idea that it would allow you to see how/when a feud started and grew over several weeks until it hit a peak and then started decreasing in frequency. In theory, this seemed ilke a good idea. But in actuality, it was way too “busy”. A good statistic is a clear, concise measure of something that is easily understandable. The Frequent Opponents stat wasn’t quite it.

So I went back to the drawing board. And instead of looking to traditional baseball statistics like battling average and home runs, I looked at the more advanced metrics that have become popular in recent years. In particular, I looked at WAR, which stands for Wins Above Replacement. Like many other stats, it attempts to quantify a player’s various contributions, whether they be at the plate, on the bases, or in the field. But what is unique about WAR is how it is expressed. It attempts to quantify exactly how many victories an individual player was responsible for. It does not actually measure that; it’s not how many times they hit a game-winning home run or made an amazing play in the field that, if they didn’t make that play, would have resulted in them losing the game. It’s merely taking all of their achievements in aggregate and answering the question “if this player was not on the team, and instead you had a replacement-level player (i.e. a below-average major league caliber player), how many less games would the team have won?”

Thus, WAR approximates something that is easily understandable. We see exactly how much one player meant to a team, how much better they are for him being on their team measured in wins and losses.

I wanted to come up with a something similar here. And eventually I settled on FLW, which approximates something that is easily understandable, how long a feud lasted in a territory. And just like WAR, it isn’t an actual measurement of that thing, but it’s a reasonable estimate. The FLW for Lawler vs Dundee was 6.72. What this means is that if you take into account how many matches they had in all the various towns in the territory over a defined period of time, in aggregate it lasted slightly less than 7 straight weeks (11 in Memphis, 8 in Louisville and Evansville, and between 1-4 times in a few other towns while not happening at all in others).

For another example of this, let’s look at the feud between the Assassins (Tom Renesto & Jody Hamilton) and the Kentuckians (Grizzly Smith & Luke Brown) in Leroy McGuirk’s territory in the first quarter of 1966. This feud was booked for four straight weeks in Tulsa (Mondays), Little Rock (Tuesdays), Wichita Falls (Thursdays), and Oklahoma City (Fridays). It ran for three non-consecutive Wednesdays in Springfield. It ran two straight weeks in Shreveport, and those shows were held on Sundays instead of the town’s regular night of Monday (this way they could run the match in both Tulsa and Shreveport). The match did not happen on the Saturday shows in Joplin (which was a smaller town and thus had a smaller budget, plus they did live TV in Oklahoma City on Saturday nights and one or both of the teams were probably there each week). So it ran for four weeks in four towns, three weeks in another, two weeks in yet another (on a day where they only ran shows those two weeks), and zero weeks in the other town we have records for. So we would expect the FLW to be a number less than four but likely above three. And when we calculate the FLW for this feud, it comes out to 3.57, pretty much what we would expect.

How do we define a feud? Are there exceptions to this definition?

The FLW stat is based on this definition of a feud: “a series of matches between the same two wrestlers or teams, occurring repeatedly in a short period of time, usually with rematches in some or all cities that often include escalating stipulations.”

This is a resaonably accurate definition of a feud, for most territories, most of the time. But there are exceptions. Here are two examples:

In 1976, a feud between Bill Watts and Terry Funk in Leroy McGuirk’s territory specifically built to one match, which was held at the first wrestling show at the Superdome in New Orleans. Watts beat Terry Funk in a match that aired on TV shortly before Funk won the World Heavyweight title (December 1975). This of course led to Watts wanting a shot at the World title, since he held a victory over the new champ. They ran a storyline in which Funk refused to defend the title against Watts. It should be noted that Funk did face Watts twice in the territory during the time this angle played out, once in Lafayette and once in Shreveport. But for the most part, the “feud” was built around Funk avoiding a match with Watts. Funk eventually put a bounty on the Cowboy. This was done to lengthen the feud and create demand for the match, with Watts fending off various bounty hunters, all the while hoping that Funk or the NWA would finally grant him a shot. And the match was finally held on July 17 at the Superdome, where it was one of two featured bouts (the other being a “Jim Bowie death match” between Dick Murdoch and Killer Karl Kox) that drew a reported 17,000 fans (in a town where they usually ran a 3,800-seat building weekly). So by our definition of a feud, could you consider this a feud? No. But it clearly is a feud.

Or how about the feud between the Midnight Express and Bill Watts in 1984? This was done through a couple of angles on Mid-South TV, and it was specifically designed to be a one-and-done match (with a masked JYD as Stagger Lee teaming with Watts to face the Midnights) in every market. It occurred 17 times in 17 different cities over the course of six weeks. This is different from the Watts-Funk feud; they’re both theoretically a “one match” feud, but the Watts-Midnights feud played itself out once in each individual TV market. We know for a fact that it set box office records in almost every town it occurred, so by any reasonable measure it was a hugely successful feud. But as far as using the FLW Score to measure it, we find that it has a lower FLW than the Midnight’s feud with the Rock & Roll Express that same year (which lasted longer and had two separate iterations). So again, this is a rare exception to the rule. However, it’s worth noting that even though each individual Midnights-Rock & Roll match drew less fans on average than the Watts & JYD-Midnights matches, it drew more money overall since it happened so many more times in main events.

What factors can affect the FLW Score?

FLW is great for approximating the length of feuds between wrestlers who are wrestling most every night that the territory is running shows. But if one or more of the wrestlers involved are wrestling less frequently than the rest of the crew, this would affect the FLW. While we’ve all heard wrestlers from the era talk about how they all worked seven nights a week and twice on Sundays every week (and Hulk Hogan somehow claiming that he wrestled 400 nights per year), in reality this is not true. There are two main groups of wrestlers who worked less frequently than the bulk of the crew (in most territories at most times). The first category is preliminary wrestlers, in particular the younger ones. There is significant evidence that they are booked to wrestle less often than the wrestlers higher up on the cards. In many cases, they’ll work as referees when they’re not in a match, or at the very least they’re at the shows to observe and learn and perhaps to be available as a fill-in for an injured wrestler. The other group is your top-level superstars, in particular the ones that also own and/or book the territory. Examples of this include Fritz Von Erich, Verne Gagne, Eddie Graham, the Sheik, and Bill Watts. In addition, the babyface WWWF champions (Bruno, Pedro, Backlund, and Hogan at times) often worked a lighter schedule than the rest of the crew. They typically work only the larger cities on the circuit. If, for example, Bill Watts is only working three nights a week in a territory that runs six nights per week, he would have to have about twice as many matches against the same opponent to have the same FLW as guys that are wrestling each other six nights per week.

The FLW Score can also be affected by the overall booking philosophy of the territory. Some territories ran the same exact matches night after night in the same week (this was often due to how their television was distributed) while others spread it out over a several week period, with a feud starting in Memphis (or Oklahoma City or Tampa) a week or more before it began in other cities in the territory. Territories with a high turnover may burn through matches in a shorter period of time, whereas a slower-paced territory with lower turnover (think the AWA or Central States) may spread feuds out over a longer period of time. Territories where the local promoters had more input into the booking of their cards (Don Slatton and Gory Guerrero in West Texas, Fred Ward in the lower half of Georgia) may see different feuds occurring in different cities, as opposed to a unified “Lawler vs Dundee” type feud that goes almost everywhere. And some territories liked to lengthen feuds by incorporating tag team matches involving wrestlers who are not a regular team. If Jack Brisco is feuding with Tarzan Tyler and Louie Tillet is feuding with Boris Malenko, they may stick a tag match involving all four in the middle of a series of singles matches between them.

For these reasons, the best way to look at the FLW Scores is to compare them to other scores in the same territory at the same general time frame. What’s considered a good FLW in the McGuirk territory in 1964 may be a low FLW in 1980. And a high FLW in Central States in 1974 may be a low number for the Gulas territory at the same time.

Another major factor affecting the FLW is how long a period we measure. For the most part, I will show FLW scores for a three-month period of time, using strict calendar quarters (January through March, April through June, etc.). Not every feud begins and ends at the same time. A feud may start in early March and last through mid-May. So it will have a low FLW for the first quarter and a medium FLW for the second while another feud in the same territory that started in April and ended in late June would have a high FLW for the second quarter. And of course, some feuds take more than three months to work their way through the territory. Or a promotion might circle back around to feud a couple of months later (like the Rock & Roll vs Midnights feud in Mid-South in 1984 or the Dick Murdoch vs Killer Karl Kox feud in the McGuirk territory that began in ’75 and had a few different iterations through late ’77). Some feuds may repeat themselves in different territories at different times (Ray Candy feuded with the Assassin in three different territories at different times). And some feuds never end, like Bobo Brazil vs the Sheik.

One factor that does NOT affect FLW is how complete our records are, provided we have a reasonable amount of data for the territory (ideally at least half, but more is always better). The one provision here is that if we have incomplete data, it needs to be a representative sample of all the data. If a territory ran two shows per night every night with one of them always being a very clear “A town” and the other clearly being a “B town”, our sample should have a representative mix of those different towns; if we only had the A towns or only the B towns, the data would be skewed.

Going back to the example at the beginning of this post, let’s say we have a territory that we know ran one show per night six nights per week. If we only have data for three of the six towns. FLW uses the number of shows we have as the denominator, so it should have very similar output even with the smaller sample size. So if they were running the same main event every night that week, even though we only have three of the shows, the FLW would be 1.00 since it was 3 main events out of 3 shows.

Much like the SPOT Rating, the FLW Scores use a rolling, weighted five-week period for many of the calculations. This helps take into account time periods where we have less shows (which could be because we are missing some shows or because the promotion ran less shows temporarily, such as around the end of the year holidays).

Is FLW perfect? Nope. And you will never hear me claim it is. I’ve mentioned before that any attempt to apply statistics to pro wrestling has its problems, because you’re not supposed to apply stats to it. That’s why the territories never truly tracked wins and losses, because a preliminary babyface is probably going to have a higher win/loss percentage than a main-event heel (historically, babyfaces win 60% of the time; also keep in mind that most prelim matches are between two preliminary wrestlers while most main events are between two main-eventers, there’s not a “balanced schedule” like in team sports where the Yankees play numerous games against good teams like the Red Sox as well as poor teams like the Orioles).

My goal with Charting the Territories is to try and use statistics to present a snapshot of what was going on in a territory in a given period of time. Who were the top babyfaces? Who were the top heels? Who were lower on the cards? Who was in the process of being “pushed”? The SPOT Ratings answer these questions. And the FLW Score answers questions like who was feuding with who and which feuds were more successful. Since we will never have complete attendance/gate records for the territories, this is a way of answering the question of “what feuds drew the best” even without that information. Based on the principle that if it drew well, the promoter would run it again. And again. And again. Until it didn’t.

Popular posts from this blog

A Dog and a King: Looking at The Junkyard Dog's experiences in the wake of desegregation and assassination