Check out all of the details of this month's Patch Notes, featuring the 16th Anniversary and VIP Renewal Update! https://mabinogi.nexon.net/news/90098/16th-anniversary-and-vip-renewal-patch-notes-march-14th
[NEW MILLETIANS] Please note that all new forum users have to be approved before posting. This process can take up to 24 hours, and we appreciate your patience.
If this is your first visit, be sure to check out the Nexon Forums Code of Conduct. You have to register before you can post, so you can log in or create a forum name above to proceed. Thank you for your visit!

Important PSA: 100% repair is (kinda) VERY GOOD!

Comments

  • IyasenuIyasenu
    Mabinogi Rep: 24,265
    Posts: 2,887
    Member
    At the very least, I'll never have to worry about my 1g base-repair costing items from ever dying.
  • BlortadBlortad
    Mabinogi Rep: 2,490
    Posts: 251
    Member
    edited July 21, 2017
    Also, I think the math is flawed in the assumption that the fail rate % and the number of times an individual has repaired a specific weapon are related, and that it's individualized in the first place. Take special upgrades for example. With those percentages, there should be a lot more successes for the individual if the math was as simple as that. I assume that the math for repairs is more complex than that "1/50 fails". I think that it's based on everyone in the game who is doing or has done repairs at that npc on that day.

    And for the record, I have very bad luck in this game, so I'm lucky with repairs, I say bout time...

    It differs from person to person because its a probability, not fixed intervals. This is literally elementary school math, 98% means that 98 times out of 100, the thing will happen. In our case its the opposite, since the rate of SUCCESS is 98%, the rate of FAILURE is 2%. This means it will happen 2 times out of 100. If something that happens 2 times in 100 attempts, that means in order for it to happen 10 times requires 500 attempts. I have no idea where you got 1/50 from. If you mean the 50% chance for special upgrades from your example, that would be 50/100 or 1/2. That means that 1 in every 2 attempts will be a failure. If you are referring to something from MY numbers, it would be 1/49 for 98%, not 1/50. You are fundamentally misunderstanding the entire concept behind statistical probability.

    Also, I never said the fail rate had anything to do with the number of times you have repaired before apart from that if you repair 500 points at 98%, statistically, 10 of them will fail. I never said that the fail rate changes in any way relative to the repair count, I simply stated what the statistical probability dictated for a given number of points, 500, based on a given percentage, 98%. 98% of 500 is 490, 2% of 500 is 10. These are facts. Go grab a calculator, seriously, because thats what I did. Enter 0.98x500, then 0.02x500. Tell me if you dont get 490, followed by 2. 98% success across 500 points does, in fact, make for 10 lost points.
  • clonedstarsclonedstars
    Mabinogi Rep: 640
    Posts: 14
    Member
    edited July 21, 2017
    Took the liberty of doing this. Left side is blessed, right side is not blessed. Also calculated costs for a more expensive weapon. Note that I dunno the exact cost of 100% repair for the weapons so I just doubled the 98% repair rate one. Should be roughly there. tl;dr it's barely worth it for a relatively cheap CRK if repairing without blessings, let alone anything with higher repair costs which is very easily found in this day and age in mabi.
    RDDuKCb.png
  • ShaeliShaeli
    Mabinogi Rep: 3,430
    Posts: 359
    Member
    edited July 21, 2017
    Being a damage junkie is expensive.
    At least I can go on using my second-rate weapons to clear out trash mobs in everyday content and not have to worry about replacing them every 2-3 months.
  • BlortadBlortad
    Mabinogi Rep: 2,490
    Posts: 251
    Member
    Ok, here, if you take a 15 dura weapon and break and repair it over and over again until you have paid for 500 points worth of repairs at the blacksmith it takes 34 repairs from 0 to 15. Any time the max durability goes down during any of these repair attempts, it increases the number of times the weapon must be repaired from 0 to reach the 500 mark. This means that the actual number of times you will have to repair the weapon from 0 to full will actually be much more then 34. Statistically, every 3-4 full repairs, assuming 15 points are being paid for on that individual repair, results in a failure. So after 3 sets of 15, you recalculate, 500 total repair point goal for the calculation - (15max dura x 3full repairs) = 455 to go to reach 500 points repaired. 455/14(the new max dura after 1 fail) = 33. (notice how we just gained 2 additional remaining repair count, we lost 3 off the 34 but are now still only at 33 because the amount that makes a full repair has gone down) 14 points per repair is still 3-4 repairs to a fail, but getting closer to 4. After 4 more full repairs, you lose 1 more point according to statistical probability, again, your specific experience will be slightly different because probability is not fixed. Now you calculate 455 - (14 x 4) = 399 You are now at 13, so the remaining 399 will take 31. This time, we also gained 2 on the count of repairs from 0 to full for 4 total additional rounds of repair needed to reach 500 as a result of dura loss, 33-4=29, 29+2=31. This time its a pretty solid 4 full repair attempts to trip that 1/49 failure rate. This means 399 - (13 x 4) = 347. 347/12=29, another -4 then +2, so our additional repair count, the number of repairs past the initial 34 that will be required tor each 500 as a result of dura loss, is now 6. We have now reached another threshhold, barely, to hit 49 or over according to probability we need now 5 full repair attempts. 347 - (5x12) = 287. 287/12=24. Because of how close 12x4 is to 49, we dont get any extra numbers this time to add to that 6, we subtracted 5 from 29 but didnt end up adding back to it when we recalculated for current dura, it remained at 24. 287 - (5 x 11) = 232. 232/11=21. This time, we did 5 full repairs but only went from 24 to 21, meaning that 6 is now an 8 for the extra full repair count. I am going to continue this to the end, if you're with me so far, feel free to skip ahead because its much the same the rest of the way. 232 - (5 x 10) = 182. 182/10=19. (remember, that 19 is number of times you repair from 0 to full) We did 5 repairs, but are now at 19 from 21, meaning 3 were gained this time, 3+8=12. 182 - (6 x 9) = 128. 128/9=14. Did 6 but only went down by 5, so 12 is now 13. 128 - (7 x 8) = 72. 72/9=8. Used 7 but went down by 6, 13 is now 14. At this point, the weapon has 6 dura left so we can stop here just fine for the purpose of the example since only 1 more fail, which will take 9 more full repair attempts roughly. The number 14 is the number of extra full repairs on top of the 34 that is the number of repairs required to reach a total of 500 at 15 max dura, so 34+14=48 repairs from 0 to full on a 15 max dura weapon to reduce the max dura to 5. The number 14 will change depending on where in this process the fails ACTUALLY happen, compared to where they statistically SHOULD happen, so its not exact, probability never is, thats why its probability, but that means in order to get a weapon from 15 max dura down to 5 max dura using 98% repairs you would have to repair it from 0 to full an average of about 48 times. Since its a total of 10 points lost, that means 10 out of your 48 full repairs will fail 1 single point. If they happen to occur directly on their statistical mark, you will get 48 full repairs from 0 before your weapon needs replacement from dura loss. Idk about anyone else, but generally, when I use a roughly 15 dura weapon for all content I only need to repair it once or twice per week. which means it would take ME roughly 48 weeks, or just short of an entire year playing for 4 hours or more every single day. Given that I only play about 2/3 of every day and that I dont generally use a single weapon consistantly for all content, only usually about half or so of it, I change it up, those 48 full repairs will last me well over a year, maybe even over a year and a half. Thats with using a weapon for general use too, its even longer if you use it sparingly. This lines up perfectly with the decay rate of my Dowras. I had them for just over a year before they were ground to nubs because I used them exclusively that entire time as I had just gotten master gunner and fancy new guns to go with it.

    If you want proof Im right, read stuff, do it. If you just want to be lazy, take my word for it when I say that the decay rate of weapons, exactly as I calculated, means a given weapon will last more then a year at 98% even if its your general use weapon, even longer if you use it sparingly. This lines up just fine with the decay rate everyone is experiencing, nothing anyone has discribed so far goes outside an acceptable range for statistical probability. My numbers discribe the decay rate perfectly, you are simply underestimating just how much 500 points worth of repairs actually is and how long it takes to repair that many times.

    If you want to try to claim my math is wrong, do it with math, not with "I dont think so" magic and baseless claims of things like adjusted probability based on the number of visitors to that npc that day. If you do not have math to back up such a claim or math that proves mine wrong, then you are, in fact, wrong. Facts are facts, numbers are numbers, math is math. If you think Im wrong, check my math yourself BEFORE trying to call me out, because trust me, I have checked my own math, as far as the calculations in the original post, time and time again by this point because I keep needing to refrence it.
  • BlortadBlortad
    Mabinogi Rep: 2,490
    Posts: 251
    Member
    Took the liberty of doing this. Left side is blessed, right side is not blessed. Also calculated costs for a more expensive weapon. Note that I dunno the exact cost of 100% repair for the weapons so I just doubled the 98% repair rate one. Should be roughly there. tl;dr it's barely worth it for a relatively cheap CRK if repairing without blessings, let alone anything with higher repair costs which is very easily found in this day and age in mabi.
    RDDuKCb.png

    Im sorry, but 99% does not equal 1/1000, it equals 1/100. Here is a link to the very wiki page that chart claims to be refrencing to get the 1/1000.

    http://wiki.mabinogiworld.com/view/Blessing

    It clearly says 99%, and nowhere does it say or even so much as slightly suggest its 99.9% and not just flat 99%.
  • clonedstarsclonedstars
    Mabinogi Rep: 640
    Posts: 14
    Member
    I gave you my math so I kinda need clarification on who you're talking to right now.
  • BlortadBlortad
    Mabinogi Rep: 2,490
    Posts: 251
    Member
    I gave you my math so I kinda need clarification on who you're talking to right now.

    The one that quoted you was directed at you. 99% success, or 1% fail, does not make for 1/1000, only 1/100.
  • clonedstarsclonedstars
    Mabinogi Rep: 640
    Posts: 14
    Member
    Blortad wrote: »
    I gave you my math so I kinda need clarification on who you're talking to right now.

    The one that quoted you was directed at you. 99% success, or 1% fail, does not make for 1/1000, only 1/100.

    I don't see where you got that from. Do tell.
  • JJJJ
    Mabinogi Rep: 5,400
    Posts: 500
    Member
    edited July 21, 2017
    Blortad wrote: »
    I gave you my math so I kinda need clarification on who you're talking to right now.

    The one that quoted you was directed at you. 99% success, or 1% fail, does not make for 1/1000, only 1/100.

    The asterisk saying, "*Repair fails every 1000[...]" is an extension of the title, "[Number] of fails expected *".
    Chart shows about 10 fails. 10 fails in 1000 attempts is 1/100, or 99% success rate. Hope this clarifies :)
  • CrimsọnCrimsọn
    Mabinogi Rep: 65,165
    Posts: 9,158
    Member
    Blortad wrote: »
    Ok, here, if you take a 15 dura weapon and break and repair it over and over again until you have paid for 500 points worth of repairs at the blacksmith it takes 34 repairs from 0 to 15. Any time the max durability goes down during any of these repair attempts, it increases the number of times the weapon must be repaired from 0 to reach the 500 mark. This means that the actual number of times you will have to repair the weapon from 0 to full will actually be much more then 34. Statistically, every 3-4 full repairs, assuming 15 points are being paid for on that individual repair, results in a failure. So after 3 sets of 15, you recalculate, 500 total repair point goal for the calculation - (15max dura x 3full repairs) = 455 to go to reach 500 points repaired. 455/14(the new max dura after 1 fail) = 33. (notice how we just gained 2 additional remaining repair count, we lost 3 off the 34 but are now still only at 33 because the amount that makes a full repair has gone down) 14 points per repair is still 3-4 repairs to a fail, but getting closer to 4. After 4 more full repairs, you lose 1 more point according to statistical probability, again, your specific experience will be slightly different because probability is not fixed. Now you calculate 455 - (14 x 4) = 399 You are now at 13, so the remaining 399 will take 31. This time, we also gained 2 on the count of repairs from 0 to full for 4 total additional rounds of repair needed to reach 500 as a result of dura loss, 33-4=29, 29+2=31. This time its a pretty solid 4 full repair attempts to trip that 1/49 failure rate. This means 399 - (13 x 4) = 347. 347/12=29, another -4 then +2, so our additional repair count, the number of repairs past the initial 34 that will be required tor each 500 as a result of dura loss, is now 6. We have now reached another threshhold, barely, to hit 49 or over according to probability we need now 5 full repair attempts. 347 - (5x12) = 287. 287/12=24. Because of how close 12x4 is to 49, we dont get any extra numbers this time to add to that 6, we subtracted 5 from 29 but didnt end up adding back to it when we recalculated for current dura, it remained at 24. 287 - (5 x 11) = 232. 232/11=21. This time, we did 5 full repairs but only went from 24 to 21, meaning that 6 is now an 8 for the extra full repair count. I am going to continue this to the end, if you're with me so far, feel free to skip ahead because its much the same the rest of the way. 232 - (5 x 10) = 182. 182/10=19. (remember, that 19 is number of times you repair from 0 to full) We did 5 repairs, but are now at 19 from 21, meaning 3 were gained this time, 3+8=12. 182 - (6 x 9) = 128. 128/9=14. Did 6 but only went down by 5, so 12 is now 13. 128 - (7 x 8) = 72. 72/9=8. Used 7 but went down by 6, 13 is now 14. At this point, the weapon has 6 dura left so we can stop here just fine for the purpose of the example since only 1 more fail, which will take 9 more full repair attempts roughly. The number 14 is the number of extra full repairs on top of the 34 that is the number of repairs required to reach a total of 500 at 15 max dura, so 34+14=48 repairs from 0 to full on a 15 max dura weapon to reduce the max dura to 5. The number 14 will change depending on where in this process the fails ACTUALLY happen, compared to where they statistically SHOULD happen, so its not exact, probability never is, thats why its probability, but that means in order to get a weapon from 15 max dura down to 5 max dura using 98% repairs you would have to repair it from 0 to full an average of about 48 times. Since its a total of 10 points lost, that means 10 out of your 48 full repairs will fail 1 single point. If they happen to occur directly on their statistical mark, you will get 48 full repairs from 0 before your weapon needs replacement from dura loss. Idk about anyone else, but generally, when I use a roughly 15 dura weapon for all content I only need to repair it once or twice per week. which means it would take ME roughly 48 weeks, or just short of an entire year playing for 4 hours or more every single day. Given that I only play about 2/3 of every day and that I dont generally use a single weapon consistantly for all content, only usually about half or so of it, I change it up, those 48 full repairs will last me well over a year, maybe even over a year and a half. Thats with using a weapon for general use too, its even longer if you use it sparingly. This lines up perfectly with the decay rate of my Dowras. I had them for just over a year before they were ground to nubs because I used them exclusively that entire time as I had just gotten master gunner and fancy new guns to go with it.

    If you want proof Im right, read stuff, do it. If you just want to be lazy, take my word for it when I say that the decay rate of weapons, exactly as I calculated, means a given weapon will last more then a year at 98% even if its your general use weapon, even longer if you use it sparingly. This lines up just fine with the decay rate everyone is experiencing, nothing anyone has discribed so far goes outside an acceptable range for statistical probability. My numbers discribe the decay rate perfectly, you are simply underestimating just how much 500 points worth of repairs actually is and how long it takes to repair that many times.

    If you want to try to claim my math is wrong, do it with math, not with "I dont think so" magic and baseless claims of things like adjusted probability based on the number of visitors to that npc that day. If you do not have math to back up such a claim or math that proves mine wrong, then you are, in fact, wrong. Facts are facts, numbers are numbers, math is math. If you think Im wrong, check my math yourself BEFORE trying to call me out, because trust me, I have checked my own math, as far as the calculations in the original post, time and time again by this point because I keep needing to refrence it.

    We have a savant here.
  • BlortadBlortad
    Mabinogi Rep: 2,490
    Posts: 251
    Member
    Blortad wrote: »
    I gave you my math so I kinda need clarification on who you're talking to right now.

    The one that quoted you was directed at you. 99% success, or 1% fail, does not make for 1/1000, only 1/100.

    I don't see where you got that from. Do tell.

    Ok, I see, I misunderstood what one of the footnote lines meant. Your saying its 10/1000 at 99%, which is right. What I DONT see though, is where you accounted for the cost of getting and upgrading a new one when the old one breaks too many times to be reasonably used anymore. If you look over my numbers again you will see that I added an estimation for the cost of obtaining and upgrading a new crk from clean to replace the one that will eventually become broken until Its unusable.
  • BlortadBlortad
    Mabinogi Rep: 2,490
    Posts: 251
    Member
    Gaea wrote: »
    We have a savant here.

    Im going to assume thats sarcasm, just more of this...

    https://en.wikipedia.org/wiki/Ad_hominem

  • clonedstarsclonedstars
    Mabinogi Rep: 640
    Posts: 14
    Member
    Blortad wrote: »
    Blortad wrote: »
    I gave you my math so I kinda need clarification on who you're talking to right now.

    The one that quoted you was directed at you. 99% success, or 1% fail, does not make for 1/1000, only 1/100.

    I don't see where you got that from. Do tell.

    Ok, I see, I misunderstood what one of the footnote lines meant. Your saying its 10/1000 at 99%, which is right. What I DONT see though, is where you accounted for the cost of getting and upgrading a new one when the old one breaks too many times to be reasonably used anymore. If you look over my numbers again you will see that I added an estimation for the cost of obtaining and upgrading a new crk from clean to replace the one that will eventually become broken until Its unusable.

    Why would you have to when the money you save is more than enough to buy a platinum hammer/crystal hammer/Ferghus hammer instead? 2.1m you save from a blessed 97% is more than enough to buy any of those in the current economy. Not to mention that when boxes come out with repair protection potions, people have sold those for 500k in Alexina. Bring that repair rate up higher and you save even more. So my point still stands that with weapons such as CRK or anything higher, it isn't worth it to shell out that much money for a 100% repair. It is by far more economical to go for lower rates and fix it back up. The ONLY times I could think of to go for 100% repair is if you cannot find plat hammers, which I have been finding all over the place, or if you want to keep the durability above 20, which is simply nitpicking at that point anyways.
    ShouK
  • RalikinRalikin
    Mabinogi Rep: 815
    Posts: 17
    Member
    Blortad wrote: »
    I gave you my math so I kinda need clarification on who you're talking to right now.

    The one that quoted you was directed at you. 99% success, or 1% fail, does not make for 1/1000, only 1/100.

    I don't see where you got that from. Do tell.

    What Blor is talking about mainly references the origins of the % symbol. It's the mathematical form of the word "percent", splitting that up gives "per cent". Now we should all know cent = century, and one of those is 100 years long. Let's theoretically put this into that terminology; 15 times per century = 15%, 67 boys per century = 67%.

    Moving it back to this, what you're saying is it will occur once per millenium (that's 1000 years), not per century (100), hence your figures provide 0.1%.
  • clonedstarsclonedstars
    Mabinogi Rep: 640
    Posts: 14
    Member
    Ralikin wrote: »

    What Blor is talking about mainly references the origins of the % symbol. It's the mathematical form of the word "percent", splitting that up gives "per cent". Now we should all know cent = century, and one of those is 100 years long. Let's theoretically put this into that terminology; 15 times per century = 15%, 67 boys per century = 67%.

    Moving it back to this, what you're saying is it will occur once per millenium (that's 1000 years), not per century (100), hence your figures provide 0.1%.

    I am quite aware what % means? But where have I stated that any repair is 0.1%? The above guys have cleared up the confusion no problem so I don't understand what you're going on about now.
  • RalikinRalikin
    Mabinogi Rep: 815
    Posts: 17
    Member
    Ralikin wrote: »

    What Blor is talking about mainly references the origins of the % symbol. It's the mathematical form of the word "percent", splitting that up gives "per cent". Now we should all know cent = century, and one of those is 100 years long. Let's theoretically put this into that terminology; 15 times per century = 15%, 67 boys per century = 67%.

    Moving it back to this, what you're saying is it will occur once per millenium (that's 1000 years), not per century (100), hence your figures provide 0.1%.

    I am quite aware what % means? But where have I stated that any repair is 0.1%? The above guys have cleared up the confusion no problem so I don't understand what you're going on about now.

    Well, it appears I was late to the party. My apologies.
  • BlortadBlortad
    Mabinogi Rep: 2,490
    Posts: 251
    Member
    Blortad wrote: »
    Blortad wrote: »
    I gave you my math so I kinda need clarification on who you're talking to right now.

    The one that quoted you was directed at you. 99% success, or 1% fail, does not make for 1/1000, only 1/100.

    I don't see where you got that from. Do tell.

    Ok, I see, I misunderstood what one of the footnote lines meant. Your saying its 10/1000 at 99%, which is right. What I DONT see though, is where you accounted for the cost of getting and upgrading a new one when the old one breaks too many times to be reasonably used anymore. If you look over my numbers again you will see that I added an estimation for the cost of obtaining and upgrading a new crk from clean to replace the one that will eventually become broken until Its unusable.

    Why would you have to when the money you save is more than enough to buy a platinum hammer/crystal hammer/Ferghus hammer instead? 2.1m you save from a blessed 97% is more than enough to buy any of those in the current economy. Not to mention that when boxes come out with repair protection potions, people have sold those for 500k in Alexina. Bring that repair rate up higher and you save even more. So my point still stands that with weapons such as CRK or anything higher, it isn't worth it to shell out that much money for a 100% repair. It is by far more economical to go for lower rates and fix it back up. The ONLY times I could think of to go for 100% repair is if you cannot find plat hammers, which I have been finding all over the place, or if you want to keep the durability above 20, which is simply nitpicking at that point anyways.

    Because over that 8.7mil figure that is the cost of 500 points of repair, you lose 10 max dura, plat hammers are 1.5-2mil each, 10 hammers is 15-20mil. You only save 8.7mil per 10 points lost by doing 98% instead of 100%, but the cost to recover those points via hammer is 15-20mil. That isnt sustainable. I guess it IS better then replacement using no hammers, but that still makes 100% better for your bank account overall. Even at 99% instead of 98%, it only even comes up the same if your hammers cost on the lower end. You would be saving 17.4mil instead of 8.7mil, but the cost of enough hammers to reverse all that would be the aforementioned 15-20mil. Also, as someone somewhere pointed out, the supply of hammers is not endless since their event only. There is no guarantee that they wont go up in price or that they will even be available at all.
  • clonedstarsclonedstars
    Mabinogi Rep: 640
    Posts: 14
    Member
    edited July 21, 2017
    Blortad wrote: »

    Because over that 8.7mil figure that is the cost of 500 points of repair, you lose 10 max dura, plat hammers are 1.5-2mil each, 10 hammers is 15-20mil. You only save 8.7mil per 10 points lost by doing 98% instead of 100%, but the cost to recover those points via hammer is 15-20mil. That isnt sustainable. I guess it IS better then replacement using no hammers, but that still makes 100% better for your bank account overall. Even at 99% instead of 98%, it only even comes up the same if your hammers cost on the lower end. You would be saving 17.4mil instead of 8.7mil, but the cost of enough hammers to reverse all that would be the aforementioned 15-20mil. Also, as someone somewhere pointed out, the supply of hammers is not endless since their event only. There is no guarantee that they wont go up in price or that they will even be available at all.

    Just follow my chart. Repairing blessed at 97% leads to saving 21.3m for repairing 1000 points, while failing 10.1 points. This means each failed point is worth roughly 2.1m. Which is, by your words, clearly enough for a plat hammer. You seem to be calculating repairing without blessings, which no smart player would actually do if the weapon is worth anything at all. This is also only considering the relatively low cost rate of CRKs. There are plenty of weapons that have repair rates at a much higher cost due to enchants.

    Yes platinum hammer supply is not infinite, but the demand is also not all that high. With blessings, VIP, music step buff, and for the CRK's case, durability loss reduction, along with final hit not taking durability, it takes quite a bit of time before you actually use a significant amount of durability. Not to mention that if you have a problem with fails, you can just use weapons that are much more disposable for most content, where failing repairs don't even matter. That reduces the repairs you would do even further.
  • BlortadBlortad
    Mabinogi Rep: 2,490
    Posts: 251
    Member
    Blortad wrote: »

    Because over that 8.7mil figure that is the cost of 500 points of repair, you lose 10 max dura, plat hammers are 1.5-2mil each, 10 hammers is 15-20mil. You only save 8.7mil per 10 points lost by doing 98% instead of 100%, but the cost to recover those points via hammer is 15-20mil. That isnt sustainable. I guess it IS better then replacement using no hammers, but that still makes 100% better for your bank account overall. Even at 99% instead of 98%, it only even comes up the same if your hammers cost on the lower end. You would be saving 17.4mil instead of 8.7mil, but the cost of enough hammers to reverse all that would be the aforementioned 15-20mil. Also, as someone somewhere pointed out, the supply of hammers is not endless since their event only. There is no guarantee that they wont go up in price or that they will even be available at all.

    Just follow my chart. Repairing blessed at 97% leads to saving 21.3m for repairing 1000 points, while failing 10.1 points. This means each failed point is worth roughly 2.1m. Which is, by your words, clearly enough for a plat hammer. You seem to be calculating repairing without blessings, which no smart player would actually do if the weapon is worth anything at all. This is also only considering the relatively low cost rate of CRKs. There are plenty of weapons that have repair rates at a much higher cost due to enchants.

    Yes platinum hammer supply is not infinite, but the demand is also not all that high. With blessings, VIP, music step buff, and for the CRK's case, durability loss reduction, along with final hit not taking durability, it takes quite a bit of time before you actually use a significant amount of durability. Not to mention that if you have a problem with fails, you can just use weapons that are much more disposable for most content, where failing repairs don't even matter. That reduces the repairs you would do even further.

    -double checks wiki- wait... So you've been able to get 99% our of 97 npcs this whole time? Jesus christ, iv wasted so much gold on edern... I thought blessing only gave +1%.

    So in conclusion then, (97% + blessing) > 100% > (98% + blessing), meaning as long as you pay attention and don't accidentally repair at 97 unblessed, that 1% difference that I was unaware of the existence of was literally what broke the math just enough for it to even out damn close to exactly, with a very slight favor toward 97%, which I didn't even give a chance as a candidate. So in effect, If you hate searching shops or things like hammers, go ahead and use 100% because its not a HUGE overall difference, but otherwise use blessed 97 and just pretend 98 doesn't exist.