I Created a Statistic; Here's Everything Wrong With It

If you've been following this blog, you know the key feature is the SPOT ratings, which measures a wrestler by their average spot on the card over a period of time. It most certainly has its limitations, many of which I will discuss in this post. But the overarching theme and what you should take away from this post is this ... SPOT is best used to evaluate wrestlers in a specific territory in a specific time frame.

What's Wrong With SPOT:

All Towns/Venues are Not Created Equal - The way the SPOT formula currently works, the wrestlers in the advertised main event of a specific card get equal billing. So the main event for a card at Madison Square Garden would get as much "credit" as the main event at the Jaffa Mosque in Altoona, PA. It's obvious that main eventing the Garden means more, but what SPOT does is merely aggregate a wrestlers' position over a variety of cards, some big and some small. I certainly could add some sort of factor that takes into consideration the size or importance of the venue, but it would create a lot more work and, in my opinion, only a minor adjustment in the results.


All Cards are Not Created Equal - Let's take the following example: a territory runs two shows in one night with a split crew. Each show has 4 matches. Town A advertises a double main event, with two matches of significance involving wrestlers who often are in the main events, plus two preliminaries. Town B has one main event, followed by a semi-final involving wrestlers that typically are in the mid-card, and two preliminaries. The SPOT formula doesn't take into consideration that the second-to-last match in Town A might be more important than the second-to-last match in Town B; they are each valued the same. As with towns and venues, to attempt to weight individual matches on a card would involve a lot more work with minimal differences to the results. It would also rely on judgement calls on my part as to how much each match should be worth, and individual biases could come in to play.


Tournaments, Battle Royals and Captains Matches - The general rule of thumb when I assign values to the matches on a card is to credit each wrestler for one match. Tournaments, battle royals, and captains matches are often handled differently.

Tournaments: if I can treat the tournament as one "match" on a card, and the card is structured similar to other cards held in that town, then I would use it. If the entire card is a one-night tournament or if there is only one non-tournament match, I would not use the card in my calculations. I also consider the breadth of the tourney participants; if it features a mix of main eventers, mid-carders and prelim wrestlers, it is less likely that I would include the card. If it features participants of relatively equal footing, it is more likely that I would include it.

Battle Royals: If the main event is a battle royal featuring all the wrestlers on the card, and the rest of the card is structured similarly to a typical card for that town, then I would not count the battle royal as a match in my calcs. One could say that having the Battle Royal advertised as the main was done because the second-to-last match is not "main event" worthy, and that's a valid point, but I currently don't consider that. If the main event is a Battle Royal featuring someone like Haystacks Calhoun or Andre the Giant, and they don't appear in another match on the card, then I would give them credit for being in the main event of that card, since it's clear they are meant to be "the draw".

Captains Matches: A lot of territories that ran split crews had cards where the main event was a tag team match, and the first one or two matches were singles bouts involving the competitors in the main. Since I generally credit wrestlers for their match highest up the card, I don't count these opening matches, but I DO use them to determine the total number of matches on the card. The more I think about it, the more I think I shouldn't do it that way, but haven't pulled the trigger on making that change yet. A wrestlers' SPOT rating for a particular card will change based on the total number of matches, so this would have the effect of lowering the scores for a lot of wrestlers.


Imbalances in the Face/Heel Structure - Generally speaking, a territory has a seet number of wrestlers capable of being in the main events, and it's split evenly between babyfaces and heels. But there may be times where they are top-heavy on one side or the other (and sometimes they might be heavy on one side by design). So some guys may temporarily find themselves lower on the cards than they normally are.


How Cards are Structured - Within a territory, different towns may have a certain formula to how the cards are structured. In particular, if the territory runs split crews several nights a week but also has 1 or 2 nights where they run one show with all (or most) of the talent, there could be a very different feel as to how those cards are put together. Another factor to consider is that this could change over time, so a .690 SPOT in 1966 might not mean the same thing as a .690 SPOT in 1979. Further more, another territory could structure their cards in a completely different manner, and a .690 SPOT in Florida in 1966 might not mean the same thing as a .690 SPOT in Amarillo in 1966. Again, the SPOT rating is best used as a way to evaluate a wrestlers spot on the card in a particular place at a particular time.

Another factor to consider in how cards are structured is the fact that tag team matches are more likely to be higher on the card, and prelim matches are generally singles bouts. This has the effect of having more wrestlers with higher SPOT ratings, since there are more "spots" available higher up the card.


The Bicycle Effect - Many territories had their TV show air in different markets in different weeks. It was expensive to produce multiple copies of one show, so after the show would air in one market, it would be sent to the next town to air at a later date. An angle based off the TV might be the main event in one town one week, and another town several weeks later. This could skew the results somewhat, as a wrestler pushed to the top via a hot angle might see a lag in when that push happens in different towns.


How to Define a "Regular" - I measure SPOT ratings for what I consider to be "regular" wrestlers in the territory. I loosely define this as someone who works at least a couple of nights a week for at least a couple of weeks. Without complete data, there can be wrestlers where I'm not quite sure if they are a regular or not, particularly among the preliminary wrestlers, who may wrestle more often on nights where there is a split crew as opposed to an all-hands-on-deck show. Additionally, prelim wrestlers are often used as referees on shows where they don't wrestle, so I can't get an accurate count of how many nights they are a part of the show in some fashion. This is yet another reason why the SPOT ratings you see are usually top-heavy, with more wrestlers clustered with higher SPOT ratings. Since many of the prelim wrestlers are not full-timers, they are not shown in the ratings.


Intentional "False Advertising" - Some territories would advertise a card in the newspaper, but would run an angle on TV leading to it changing. Tennessee in particular was known for this; they'd run an angle on TV where the babyface would demand to get his hands on the heel who wronged him, even if it meant switching the whole card around. In the long run, this probably doesn't change things too much, as whoever was put in that spot on the original advertised cards was probably deemed worthy of being in that level (main event, mid-card, etc) of a match.


Touring National Stars - When the World Heavyweight champion comes to town (or Andre, or Haystacks, or a former World champion, or someone with significant name value), they are typically in the main event, or at the very least high up on the card. This has the effect of "pushing" some wrestlers slightly further down the card, as they're taking a spot otherwise going to a full-timer. Since all cards are treated equally, a loaded card featuring the World champ plus Andre plus a quadruple main event isn't weighted differently.


Women and Midgets - I would LOVE to include SPOT rankings for women wrestlers and midgets. Unfortunately, there are two major factors as to why this is not really possible. First, they tend to come in for a week at a time (though some territories used women wrestlers that lived locally more often), so the sample size isn't really big enough. Second (and more important), these attractions were almost always placed in the middle of the card, with very little variation. I'm sure if I looked at the Fabulous Moolah over a very long period of time, she would have a SPOT higher than the other women, but over the course of a year or less, all the women (or midgets) were pretty much put somewhere in the middle of the cards wherever they went regardless of which women (or midgets) were involved.


How Long a Period of Time is Optimal? - The SPOT ratings you've seen so far are broken up by three month periods of time. It includes wrestlers who may have been there for the whole period, and ones who came or left during the time frame (with a general minimum of two weeks). So there are sample size issues in that some of the wrestlers have a larger sample. In the past, I've used a "rolling time period" where the SPOT for a wrestler at a given point in time is based on their matches for a period of time both before and after that given point. The question is how long a period of time is enough that you have a large enough sample, but not too long so that you can see where wrestlers are moved up (or down) the cards over time as their push and role changes.


There are probably a lot more points of contention to be brought up. The short version is that my current method of calculating SPOT is quite simple (if time consuming) by design. Much like the core baseball statistics that were used for years, it tells you something useful, even if it has flaws. It can be used to evaluate a wrestlers' average spot on the cards at a specific time in a specific place, and can be compared to other wrestlers in that place at that time. Beyond that, it is an imperfect tool.

One area where I do believe it has value is in evaluating the career path of a wrestler. If you track his movement over time from territory to territory and see how his spot on the cards change, it will give you a better understanding of his overall path. Of course, having a high SPOT in a small territory like Gulf Coast is a lot different than having one in a larger territory such as Mid-Atlantic, so you still would need to look at each territory differently. But it would be a neat "big picture" tool.


  1. Al, I really enjoyed your discussion of this metric on the 605 Podcast. As a person who deals with data for analytics, I find this kind of thing fascinating.

    Have you thought of addressing the Town/Venue issue by assigning a standard weighting by grouping the venues into a scale of common values based on their size/seating capacity? You could go low to high in order to get a weighted multiplier value
    5 - Large Venues (10000 or higher): MSG, the Omni, Boston Garden, Mid-South Coliseum, the Olympic, etc
    4 - High-Mid Size venues (4k-10k): Louisville Gardens, Ft. Homer Hesterly Armory, Dallas Sportatorium, Kiel Auditorium
    3 - small arenas (1k to 4k): Portland Sports Arena
    2 - Very small arenas (house shows)
    1- Spot shows

    Maybe you should have a modifier also for dedicated TV Taping arenas if they advertised cards (The Chase, Tampa Sportatorium, even studio shows)? Also maybe a special modifier for special very large cards, like Shay Stadium or the Rose Bowl

    Another thing that would be interesting is to either have another data point for gate and draw for the shows, though I would imagine this would be more fraught with error than your other data points, given the lack of reliable or even truthful records that were kept.

    1. I have thought of adjustments similar to this, as well as many others. The short answer to all of them is that it would create a LOT more work, and any improvement in the results would be (in my opinion) minimal. As I discussed this week on another podcast (http://www.classicwrestlingmemories.com), wrestlers who headline B-towns more often than similarly slotted wrestlers in the territory at that time is pertinent information. Attempting to weight the town would minimize that info.

      TV tapings aren't typically structured sequentially like house shows, in the case of places like Portland and DFW where the first hour is shown on TV and the rest are for the live crowd only, I tend to include those.

      All I'm looking to measure with SPOT is the pecking order, or hierarchy, or depth chart, for wrestlers in the same place at (relatively) the same time. Is there the possibility that I, or other (smarter) people could then go a step or two further and apply venue factors, town factors, territory factors, attendance/gate factors, etc? Absolutely!

      Currently there is a decent amount of evaluation of main eventers and tying them into houses/gates; one of the reasons I came up with SPOT was to measure EVERYBODY up and down the card.

      As mentioned in this post, it's simple by design. I like to think of it as Batting Average; it tells you something, and that something has value, but if you really dig deep you can see it has flaws and can be improved upon (but with an exponentially higher degree of number-crunching).


Post a Comment

Popular posts from this blog

Amarillo 1966 - Introduction

The McGuirk/Watts Wrestling Almanac - Chapter 1 (Q4 1959)

A guide to SPOT Ratings and FEUD Scores...featuring Magnum T.A. and The Midnight Express