More on PST's

[Joseph-Dirt]

On Steve + Jeff’s podcast with Zack the discussion about how the new PST’s came about started with the history of the Classification system and how it’s required that a “review” of the PST’s is done each year. The review is done by Jeff + Zach with involvement of a committee. The result of the reviews are submitted to the Board and they can choose to implement them or not. It was also noted that the algorithm used in the reviews has changed over the years. At some point in the podcast it was stated that it was  “too easy to make GM” in the Low Ready divisions. Editors Note: Jeff + Steve made GM under even easier standards than today. What is “too easy” and who determines that? I could not determine that from the discussion on the podcast. There was a lot of discussion about the origins of PST’s (1985 - developed for USPSA) and how the 95% standard for GM did NOT necessarily mean that the top 5% would get that classification. So what does that 95% mean? Unclear. It seems to be very subjective and open to interpretation. All that said, it was generally agreed that it’s “too easy” to make GM in the Low Ready divisions.   

 

Zack had some very interesting stats on SCSA member classifications. He asked Steve + Jeff to guess what percentage of Members across all divisions fell into the GM, M, A, B, C + D classifications. As of Dec 28, 2023 here are the numbers:

 

GM - 4%

M    - 5%

A.   - 10%

B    - 30%

C.   - 40%

D.   - 12%

 

Steve and Jeff guessed wrong for all the classes. 

 

82% of SCSA members have a classification of B or lower. We have 3 times as many D shooters as GM’s. The entire discussion was centered on what effect new PST’s have on what it takes to make GM and whether that was too easy or too hard. Not a word about what the effect the new PST’s would have on what it takes to make any other class. 

 

Zack then provided numbers as of Dec last year for the Low Ready divisions:

 

Pct of Members who are GM’s by division

 

PCCI - 15%

PCCO - 5.8%

RFPI  - 3.2%

RFPO - 3.2%

RFRI.  - 11%

RFRO - 6.1%

 

Steve + Jeff were also unaware of those numbers. 

 

I’m going to call out Jeff here. A guy who is doing the work and making the recommendations to change PST’s. It seems he’s doing that without any idea how many GM’s we have. His guess for the overall number was 14%. Then after being told the percentages for the Low Ready divisions he insisted they were all greater than 4% when RFPI + RFPO are below 4%. Is it any wonder why he pushed for such aggressive changes? 

 

Looking at the Low Ready numbers, to me it looks like PCCI and RFRI would meet a definition of “too easy” to make GM. Go ahead and crank those PST’s down. The others? Maybe PCCO + RFRO should get a tweak but the RFP divisions? They already look pretty exclusive to me. What do the new PST times do to those divisions? Drop RFPO by 4 seconds and RFPI by 8 seconds. I’m trying to imagine the reaction of an RFPI C class shooter, a couple seconds from making B class and who is totally oblivious to how any of this works (As was I when I was a C class shooter) finding out that they are now 11 or 12 seconds away from making B class. (The way the percentages work it’s not a linear increase of just 8 seconds) Are the current PST’s for the RFP divisions really that far off? How long will it be before someone makes the climb from M to GM when they have 8 more seconds to shave off their time? How is this helping our sport? 

 

Editors Note: I have never shot RFPI and have no intention of ever shooting RFPI. 

 

Going forward I would hope Zack + Jeff would provide this level of detail for all divisions when the new PST’s come out. Assuming they’ve used that data as part of their review. Which this time it appears they did not. 

[shred]

Yeah, it sure seems like there's a lot of jealousy and angst around "GM" that causes odd effects in people.

 

[GKB]

Some mitigating factors to the numbers.....

 

The majority of the Rifle Iron GMs (RFRI and PCCI) are GMs in the Optics Division, also.  I think it is more a case of easily transferred skills from RFRO/PCCO to the Irons more than those being "too easy".  Are Irons slower?  Yes, but if you can shoot GM in Optics you should be able to do the same with a little bit of work and the difference with the Iron times.

 

     RFRI has 82 of it's 94 GMs are also GMs in RFRO (87.2%)

     PCCI has 47 of it's 62 GMs are also GM in PCCO (75.8%)

 

     RFPI has 38 of it's 49 GMs are also GMs in RFPO (77.6%)

 

     PCCO has 111 of it's 134 GMs are also GM in RFRO (82.8%)

[Renno]

The class system is designed to be dynamic with shooters times and change with them. It works as a way for competitors to compare their times against similarly skilled opponents in a match. The sport is always getting faster and everyone is always getting better, so the PST has to keep adapting to spread everyone into the classes they belong. As Zack Jones noted, PSTs have never been adjusted based on a division's overall population. It's not a population distribution based on the %. It's important to understand that a division with less or more than 5% population of GMs does not indicate that the division PST is high or low. 

 

What happens when the PSTs do not change fast enough to keep up with the speed of a division? Well you end up with shooters stacking up in the higher ranks and they need to be spread apart more to keep the classification system working in matches as I described it above. This is what we see now in the low ready divisions where 95% GMs are going against up to 150% GMs. Thats not right when every other division window is about 10-20%. To fix this, PSTs need to be lowered to more accurately place shooters in the M and GM classes.

 

In this case, adding another class is not a viable solution. This inflation problem is exclusive to low ready divisions and adding another class would mess up all the centerfire divisions that have already stabilized over years. Low ready divisions have not stabilized yet because they are still getting faster. Therefore, this is a PST problem for low ready divisions, not a lack of classes.

 

Shooters often set a specific class as their goal, not understanding that the goalpost is designed to change since it's a relative skill rating. I understand how that is frustrating but that's not a valid reason to stop maintaining the balance of classifications.

 

Addressing the non-linear nature of the PSTs as you go down classifications: that's how it's intended to operate. The lower skill class you're in, the easier it is to make up time since you have a greater capacity to improve. It's much easier for a 50% C class shooter to make 16 seconds than it is for a 100% GM to make 8. A C class shooter has much more information out there to learn and push that time out. A GM can't just take a class or read a guide and shave that kind of time off where a C class shooter can. While the margin is quantitively wider for C class to B class, the changes are proportional in difficulty.

[schmidtg]

I am happy that there is finally a declared and utilized methodology for computing peak times for all divisions.

 

In theory, with enough high-skilled participants at WSSC, this means that the peak times (now derived from 95%+ times) should be well-calibrated.

 

Long-term, this should mean that the classification system in SCSA is much more well-calibrated compared to USPSA.  The latter's classification system is objectively terrible because there are almost 100 different classifiers with poor calibration (some are way too easier, and others too hard) with no consistent methodology across years (the oldest classifiers in USPSA are now from 1999) or divisions.

 

We can criticize the new peak time methodologies, but at least they are transparent and applied consistently across divisions.  That's more that can be said of any other shooting sport right now - USPSA and others use secret sauce to compute their high hit factors which yields inconsistent calibration.

 

The SCSA community has a lot to be thankful for because although the PST methodology might seem simple, it is applied in a fair and consistent manner.

[ZackJones]

A couple of quick notes here:

1 - I provided the data to Jeff and Steve because It was readily available to me.

2 - The data is 7 months old at this point. As members expire the number changes week to week. 

3 - USPSA doesn't have any kind of quota or limitation on how many can be in any class within any division. 

4 - Too easy is what I said. Purely subjective and not based on anything other than my opinion. Clearly it wasn't THAT easy because I never made it past 94.6%. 

 

[JM_]

12 hours ago, schmidtg said:

We can criticize the new peak time methodologies, but at least they are transparent and applied consistently across divisions.  That's more that can be said of any other shooting sport right now - USPSA and others use secret sauce to compute their high hit factors which yields inconsistent calibration.

 

 

keep in mind if we apply the same methodology of average of 95% GM, we will see similar time drops next year and almost zero GM's after 3-4 years.

[ZackJones]

On 7/30/2024 at 8:39 AM, JM_ said:

keep in mind if we apply the same methodology of average of 95% GM, we will see similar time drops next year and almost zero GM's after 3-4 years.

Being a GM is overrated so I hear :). 

 

[JM_]

On 8/1/2024 at 3:50 PM, ZackJones said:

Being a GM is overrated so I hear :). 

 

Thats why I'm Super GM