First Draft of Razor Aggressiveness Results

Yesterday I updated the chart on the Wiki page to reflect my current analysis of the first batch of razor aggressiveness polls. Please take a look, read below, and let me know your (positive) thoughts and (constructive) criticisms:



Buyer's Guide for Currently-Manufactured DE (Safety) Razors, Ranked by User-Polled Aggressiveness



(hosted in our Wiki at http://wiki.badgerandblade.com/Modern_Double-Edged_Safety_Razors_Ranked_by_Aggressiveness )





Some notable differences from prior rankings:

• All rankings are now determined through my makeshift formula that increases the point differences for more consistent and (relatively) higher-confidence polls. Each poll can result in a theoretical difference up to 2.5 on the scale ... although in practice the largest difference calculated was 1.7 points from the Weber > EJ89 poll.


• The razors on the mild end of the spectrum are now more condensed with greater differences between the more aggressive razors. For example, the EJ89 is now set at 2.5 instead of 3.0


• The Lord razors were removed from the list and will be re-polled in the upcoming second batch of polls that will use a different format of matching razors to some higher-confidence reference points


• Prior conclusions on the ATT razors have been reassessed with a new formula. Revised results are summarized as M2 > M1, R2~=R1, and H2 > H1

Summary of the polls:

• A series of 45 polls compared razors of similar aggression. Each poll was open for 30 days. Members cast their votes from personal experience with both razors.


• Votes were cast for Razor X, Razor Y, or Approximately Equal


• If Razor X is more aggressive than Razor Y, and Razor Y is more aggressive than Razor Z ... then Razor X is also assumed to be more aggressive than Razor Z


• None of the polls reached the number of votes that would be necessary for a statistically valid conclusion. I would probably recommend against companies making investments based on this data ... but on the other hand I think it is still more likely to be valid advice than responses to a "What 2nd razor should I buy?" post to the forum. If somebody else can draw a larger sample of shavers with experience in both razors ... then this data will be replaced.


• Presuming the data is sufficiently valid for individual purchase recommendations, we can draw conclusions that are something like: "The average Razor X is usually considered to be more aggressive than the average Razory Y, to the average shaver, with an average handle, using an average blade and average prep/technique and an average beard/skin plus with an average definition of razor aggressiveness.


• That's a lot of "average" stipulations ... so even if we had hundreds of votes per poll ... the validity of each conclusion would still depend very much on the individual shaver.

Summary of formula:

• A unexpected correlation seems to be that more consistency in the votes typically equates to a greater degree of difference in aggressiveness ranking (currently scaled up to 2.5 on the scale per poll)


• Nutshell: Difference in Ranking = ((RazorX - RazorY)/((RazorX-RazorY)/ApproximatelyEqual)) * MaxDifference * ((TotalVotes/TrustedNumberOfVotes) * ConfidenceMultiplier)
  
  
  
  
  • TrustedNumberOfVotes and ConfidenceMultiplier were adjusted until working down from the 9.5 razor through higher confidence polls ... the least aggressive razor on the list equaled 1.0 (and the previously agreed upon midpoint also conveniently ended up being 5.4 ... very close to the 5.5 true midpoint)
  
  
  • ConfidenceMultiplier ended up being 0.69
  
  
  • TrustedNumberOfVotes ended up being 15 (about twice of what we had in the average poll)
  
  
  
  

Thoughts?





Cheers,

Shawn

First Draft of Razor Aggressiveness Results