nozh2002
BANNED
- Joined
- Jun 9, 2003
- Messages
- 5,736
I like you tests, and I think that eliminating the wooden board is an easy and effective way of clarifying some test data. But still looking at the numbers, they confuse me quite a bit. How can you rate one Steel over another, when there is a big difference in results within the same knife.
Take Yuna for example, after 200 cuts the difference between 1st run (80) and 2nd run (50) is 38%. Or ELMAX after 100 cuts doing on a 1st run (45) and on 2nd run (65). That's 31% difference. After how many cuts do you establish the winner?
Yes this is an possible issue, this is why I did test Sniper Blade LPC three times to see what variations would be. All were not same but pretty close. That made me confident that in normal condition results will be stable.
Yuna and Elmax are edge cases. Yuna was first and so after year or so I decided to retest it, first run of course always different and so I did it again. With Elmax I was not satisfied with first run - I did testing late night, right after I got knife I was rushing to test this new promising steel. I did not like results and so decide to retest it next day without any rush. I decide to keep both results on the page for Yuna and for Elmax - I do not thing that editing results even in this case would be correct.
There are also cases when blade is doing better after more cuts. That is weird. I'm sorry I haven't read the whole 500 post in those 3 threads you have links to, but I have read some opinions about some weird thing like carbides alignment and blade doing better after several sharpening... I think the problem is that you're doing the sharpness test in one spot on the blade. Why not to put 5 dots and do 4 cuts on each spot, instead of 21 cuts on one spot. The CATRA machine for example doesn't have such flaw, because it does the cut/test with couple of inches on the edge.
Well we do not really understand what is edge going through. What is some carbides fall out. But I noticed this little "improvement" many times. I guess it is possible variations. I do not worry about that.
Then on top of different results from the same knife, there is a big variance within the same steel from different manufacturers. The good thing is that you're listing the exact knife in your rating. But still, how can you rate one steel better than the other when there is so much weird stuff is going on? Even rating one knife technically might be incorrect, because the same knife on a 3rd run might show different result (but that will improve statistics). Also, is there a guarantee that if I buy the same model as you have tested, it will behave exactly the same way as yours?
There is no any guarantee, you should make your up you mind yourself. But this is better at least to me then any other information available on blade performance.
At least the numbers from CATRA machine come from the same manufacturer (if I understand this correctly)? They can easily test 10 knives with the same steel and come up with average number that they can compare to the average number of the other steel (I have no clue how they actually do it).
But also that data from manufacturer's CATRA will be valid only to that particular manufacturer and their heat treatment. As everybody knows (and very nicely shown by you) same steel can be heat treated by different manufacturer very differently. More than that, even if data shows that edge is more wear resistant, it doesn't mean that steel is "better", because there is no toughness in the equation. All that it would mean that Steel A @ X HRC is more wear resistant than Steel B @ Y HRC. Without toughness it is difficult to proclaim one "better overall", than another. Better for whom and for what purpose? ZDP @ 67 HRC holds edge very long, but isn't very durable (at least from one manufacturer). That actually might be why manufacturers don't want to disclose CATRA test results. They have to find a spot where wear resistance, toughness, ease of sharpening would appeal to the public. The CATRA data alone might not speak in their favor.
I wish Spyderco post CATRA numbers for their Mules. That would be pretty awesome.
Existance of CATRA tests useless until results is published. So far this is biggest secret of knife industry and nobody want to share this. So it may be thousands times better, but if nobody know about it - what the point?
Spyderco is not going to publish results - Sal made this pretty clear to my direct request.
I think your tests are generally good, but they are missing more statistics to be really accepted. I understand that you cannot test the same knife 10 times, or better yet 10 different knives of the same model, or 100 knives with the same steel from different 10 different manufacturers, but without it I just don't know how anybody can say that D2 (replace the name) is better than S30V (replace the name) or vice versa. In some way you might even say that very subjective opinion of 100 owners from some use and subjective comparison, might be a better indicator than a result of just one of your(or somebody else) tests. Another problem of course is that those 100 subjective users might not have the same reference point . I wish there be more people trying to do something like you, but I sure am way too lazy to do it and prefer to spend my time on something else . Sorry . I appreciate your efforts though.
Yes, I also would like more tests done to have better statistical representation. More way of testing etc... Initially I was thinking some enthusiasts will join this effort and we will have some kind of community project like in programmers world - this is why I was working on making this method simple and clear so any one can do. But knife world is different. And it is only me doing this tests and I did quite a lot already, I think.
Thanks, Vassili.