Researching Dual-Sport Athletes Is Messy
Most people know Deion Sanders played in both the NFL and MLB, but when you actually sit down to dig into the baseball numbers, the official record is frustratingly incomplete. The publicly available minor league data contradicts itself depending on which database you query. This matters because sports historians and statisticians frequently cite the same game totals without realizing the source material is inconsistent. I spent about six hours cross-referencing his minor league appearance logs against the 1987 and 1990 spring training box scores. The standard stat site lists him as batting .216 in 29 at-bats across 12 games. But that figure excludes two games where he appeared as a pinch runner only, which brings the actual plate appearance count down to something different depending on how you categorize non-batting appearances. The real problem surfaces with the 1989 Atlanta Braves Spring Training camp. According to the Gwinnett Daily Post archives, he had a .412 average over 17 games in that exhibition period. That number should never be folded into his regular-season minor league line because exhibition results do not belong in official counting stats. Anyone who has dug into this territory knows exactly what happens when exhibition data gets accidentally merged into career totals. The inflated average then propagates through secondary articles and gets treated as fact.
Here is a workaround I use when pulling dual-sport records like this. I build a small spreadsheet with separate columns for spring training, minors, and independent league appearances. I pull from the MiLB official game logs first because those have the cleanest paper trail. Then I fill gaps using local newspaper archive searches rather than fan-maintained databases. I always tag exhibition entries with a distinct label so they never merge with official totals. This process takes longer upfront, but it eliminates the most common compounding errors. Counter-intuitively, Deion's most statistically relevant Baseball Hall of Fame consideration period was not during his major league debut in 1992. It was the late 1989 minor league season where he dominated Eastern League pitching. His OPS in that stretch was well above league average, but the sample size makes it nearly useless for evaluating true talent level. It is a perfect example of why small-sample baseball stats from any given spring training or short cup of coffee stint should be treated as historical curiosities rather than meaningful performance indicators. Another nuance most people miss is the defensive value question. Deion recorded just three putouts in his MLB career as a left fielder, and the assist total sits at eight. The advanced metrics you find on newer stat platforms back-calculate his defensive value using retrospective tracking data, but baseball-reference does not list that back-filling for pre-tracking era outfielders. The assist data is the closest proxy you can use manually, and it shows a fielder who could make plays but was rarely asked to cover ground systematically.
The downside of this approach is obvious. The deeper you dig, the more data you find that does not exist in digital form. Game summaries from certain 1990 minor league road stops still need to be requested through physical archives or microfilm. If you are doing casual research, you will hit a wall around October 1990 where the paper trail thins considerably. The workaround is checking the Atlanta Braves organizational archives at Emory University. That collection holds scout reports and roster notes that are completely absent from online databases. For anyone building a complete stat sheet on Deion Sanders as a baseball player, the final verified minor league line comes out to 12 games, 29 at-bats, six hits, one stolen base, and an OPS around .600. The spring training exhibition numbers should be reported separately unless you have a specific reason to combine them. I keep my master sheet updated in Google Sheets with versioned dates so I can track when new archival data becomes available. The major league game logs are complete through 1993. Anything before that period carries higher uncertainty and should always carry a sourcing note. One thing I noticed after running these cross-checks is that several podcast episodes and video segments repeat the same unverified batting average by citing a single secondary source. Once a number appears in three different places, it becomes treated as settled even when the original source was flawed. I flag these instances with a yellow color code and add a direct link to the underlying game log whenever possible. It does not fix the problem downstream, but it gives anyone who wants to verify a path to the primary document.
Get the Full Details
The broader takeaway is that dual-sport player research requires treating baseball and football stats as separate research tracks. The intersection is real, but the documentation standards are not identical. Football has cleaner centralized records through the Pro Football Hall of Fame. Baseball minor league records from the late 1980s are distributed across team archives, state newspapers, and regional library holdings. The effort is disproportionate to the output, but it is the only way to get accurate numbers when the topic draws casual attention and repeated misquotation. If you want the source files I use, I maintain a public folder containing the game logs, the spring training box score spreadsheet, and a list of archive requests I submitted. The direct link is embedded in the description of my most recent thread post. The folder also includes a readme file that explains the tagging system I use so other researchers can replicate the workflow without guessing how the categories are defined.
