Texas Gun Collectors Association

Readers of the first installment of The Survival Files (Spring 2019) have asked questions that focus on three specific areas: the pace at which serial numbers were recorded; the purpose of the stability index; and serial number registries.

The Pace of Recording. Some readers have pointed out that over the last fifty years, the pace at which serial numbers were recorded in the Boulton database must have accelerated rapidly after the invention of the Internet. This is of course correct. For any serial number collector, the pace of recording varies due to many factors. The Internet had huge impact, and it would not have been possible to compile over 85,000 serial numbers without it. It was not, however, the only significant factor.

The pace at which the Boulton database grew has roughly tracked the pace of sales. A major factor affecting sales in the UK was the British “100-year rule,” as discussed in the article on rarity in the Spring 2019 issue of The Texas Gun Collector. Other major factors included the 100th and 150th anniversary Civil War commemorations; growing interest in Civil War reenactment; the many books,

movies and television productions that generated interest in the Civil War; the number of large antique gun collections coming onto the market; and the growing number of auction houses that specialize in antique firearms.

Value of the Stability Index. Variation in the pace of collecting serial numbers doesn’t change the validity of the key survival statistic—the known survival rate, or KSR, as of a particular date. The pace just drives the speed at which the KSR grows over time.

We measure the pace at which numbers are recorded for three reasons. The first is to quantify how stable the data are—that is, how fast the KSR grows for each model. High stability index values indicate lower relative stability, and vice versa. The second reason is so we can compare how the stability index varies between models of revolver. To make this comparison possible, regardless of whether 200 or 200,000 of a model were produced, we compute the stability index as the average annual percentage increase in recorded serial numbers over the previous three years.

The third and perhaps most important reason is that the stability index is a good indicator of how many unrecorded specimens are still out there to find. Consider, for example, the stability index data generated so far:

ModelStability IndexTotal Estimated Production
.45 National (teatfire) 0.00% 25M
Colt WalkerM 0.38%M 1,100M
.31 Pettengill 1.81% 180
.34 Pettengill 1.85% 900
Freeman 2.12% 2,000
.44 Single Action Starr 4.26% 32,785
.44 Double Action Starr 4.44% 23,140
Rogers & Spencer 4.86% 5,800
.44 Pettengill 5.04% 3,300
.36 Double Action Starr 5.38% 3,100

These data fall into three fairly discrete ranges, and statistically confirm what most collect ors qualitatively observe from auctions, on-line sales and antique gun shows. At one end of the spectrum—below 1%—the ultra-low values for the .45 National and Colt Walker reflect the fact that very few unrecorded examples of these guns still exist. Our chances of finding a new unrecorded example to add to the database are very low indeed.

At the other end of the spectrum—above 4.0%—the very high values for the Starr, Rogers & Spencer and Pettengil army revolvers indicate that a lot of unrecorded specimens still survive. In the middle of the spectrum—in the 1.5-2.5% range—are the Freeman and the small frame Pettengill revolvers. Unrecorded specimens are much harder to find than the Starrs and Rogers & Spencers, but much easier to find than an unrecorded Colt Walker or a .45 National. From April 2016-March 2019, only two unrecorded .31 caliber Pettengills, three .34 caliber Pettengills, and 15 Freeman revolvers were located.

What drives variation in the stability index? Total production helps account for part of it, at least on the low end and high end of the stability index spectrum. But there’s no apparent correlation between the stability index and total production, so other factors must be at work. A big factor is likely collector demand, which as Flayderman points out is the number one driver of antique gun values.

We are just starting to do comparative analysis of stability index values. As data on additional models are calculated, we will report our findings in future installments of The Survival Files.

Serial Number Registries. A third issue raised by readers is the overlap in serial number databases. Overlap is widespread, and it highlights the need for single, model- specific, serial number registries into which collectors can feed their data so that there is one ‘most complete’ database for each model of gun. Ideally these databases would be available to the collecting fraternity. Regardless of where the individual registries reside—in institutions, or with individuals, or both—they would greatly improve the quality of survival rate statistics.

Winchester collectors are way ahead of the curve on this, helped by the archival resources and accessibility of the Cody Firearms Museum. The Winchester Collector periodically lists data collectors and their emails for over two dozen categories of Winchester. There are also serious Colt, Sharps, Remington and Confederate serial number collectors. Many of these databases are proprietary, and a few are commercial.

For collectors who want to record their own piece of history, and add to our knowledge about the survival rates of these great American guns, contributing to one of the serial number databases is a great way to do it. The Boulton database is the most extensive in terms of the number of models represented, and it is the only database for many of the more obscure and low production American percussion revolvers. Those with serial numbers for American percussion revolvers are urged to send them to Phil at philboultoncps@hotmail.com.