# Thread: Dodge and Mobility: 3% + 2% = 4% ??

1. I always assumed the 2% was intended to be "2% of the 3%" added and then rounded up. Not that that would make much sense either, but I also use Rest's method of stacking **** till it hits max. Plus bothering with math in DDO involves trusting that the numbers displayed and the actual calculations relate to each other, and I can't commit to that.

2. Originally Posted by Vargouille
I spoke with the engineer who looked at this bug, and this is 99.999999% only a display issue, along with a tiiiiiiny unintended "nerf".
Essentially, unintended floating point math is resulting in you actually having approximately 4.9999999% dodge instead of 5%, and the display code is truncating the number down to 4. You are in fact getting almost all of the dodge, but it is technically true that the Dodge Feat is only giving an exact value of 0.029999999 Dodge instead of 0.03. (That right there is the exact count of significant digits.)
I'm told this should be fixed in an upcoming patch.
Well, that's cool. When I first read the thread title, I presumed that the math was working the way it does in pen-and-paper D&D -- percentage increases are based on the original amount. In other words, 2%+3%+8%=11%, not 13%. I guess this is one of those differences where the video game and the pen-and-paper version differ.

3. Originally Posted by Ryiah
Except in the sense that the first two numbers require double-precision floats to retain their digits whereas the second two numbers would only require single-precision. Whether or not switching to single-precision would make any meaningful difference though is another matter.
Whatever terminology is used to describe the problem here, placement of decimal points is in practical terms a source of many bugs of this sort, and indeed in some highly critical banking, accounting, commercial or other such financial software applications, it can on occasion be better to simply ignore the decimal point, make certain calculations as if everything was a whole number (which is to say, to treat 0.029999999 as being 29999999 -- and which is BTW exactly how the CPU will treat all such numbers anyway in the machine code), and add the point afterwards by a division by 100 (or whatever appropriate multiple of 10) after the more complex calculations have all been dealt with (which would not necessarily even require any extra lines of code, if you're tricky enough).

Which would probably NOT be the appropriate fix here, not that the Dev likely needs any "help" in the first place, I'm just saying there's more than one way to skin a cat, particularly if you're seeking optimal performance ...

4. Originally Posted by Vargouille
I spoke with the engineer who looked at this bug, and this is 99.999999% only a display issue
Looking at the explanations given to you by the Dev, it's actually 99.99999% only a display issue, not 99.999999% -- you got a floating point wrong, and you have therefore understated the true scale of the problem by a factor of 9.999998.

5. While you are fixing dodge, can you please fix the dodge on the dragontouched soveriegn rune. It is still stuck at 3% even though the rune now reads at 6% and has been this way since dodge was doubled and made non stacking a while back.

Thanks.

6. Originally Posted by Cardtrick

I've definitely seen the behavior the wiki describes -- tier 1 gives 1% dodge, tier 2 increases that to 2% dodge, but tier 3 does absolutely nothing (leaves it at 2% dodge).
Maybe it acts differently at different times. According to the explanation given by Varg, that would explain it. I know my description is accurate for the behaviour I saw because I respec'd that tree three times testing it (cost me a whopping 80p each time).

7. Originally Posted by Merlin-ator
TL;DR: Floating point math is the binary version of decimal points

Floating point math is annoying. Unfortunately, certain numbers (like 0.1) are impossible to represent, similar to how decimal cannot represent 1/3 or 2/3.
Huh. Cool - thank you!

So...if aliens had math, it might be different from ours and unable to represent certain values? That's cool. Maybe we couldn't understand how they would represent numerical values.

8. Originally Posted by Coyopa
Maybe it acts differently at different times. According to the explanation given by Varg, that would explain it. I know my description is accurate for the behaviour I saw because I respec'd that tree three times testing it (cost me a whopping 80p each time).
Yeah, knowing that there's floating point math and rounding issues involved, it probably depends on what other dodge values are being added to these.

9. Originally Posted by Vargouille
I spoke with the engineer who looked at this bug, and this is 99.999999% only a display issue, along with a tiiiiiiny unintended "nerf".
That is hilarious. But the OPs post also reminded me of this.

10. Originally Posted by Missing_Minds
That is hilarious. But the OPs post also reminded me of this.

Most excellent...hadn't seen that in years.

11. Originally Posted by Nibor
BTW, this is a classic example of why the userbase feels that submitting bugs through the bug reporting tool is useless. I submitted this exact bug, along with 100% recreate description, over a year ago when the new bug reporting tool was live. It was filed into a black hole for all we can tell. Someone mentions it on the forums and it gets a reply and an explanation.

Why should I bother with submitting more bug reports? I have no reason to believe anyone looks at them, and significant evidence that no one does.
Maybe they were able to respond so quickly on the forum because they had read and investigated your bug.

12. Originally Posted by Singular
Huh. Cool - thank you!

So...if aliens had math, it might be different from ours and unable to represent certain values? That's cool. Maybe we couldn't understand how they would represent numerical values.
They could still accurately represent rational values as fractions, if not base number floating point.

13. I don't care much wheter they use floating point in double or float or convert to decimal to avoid non intuitive floating point corner cases.

But what I do care about: do *not* use truncate. Round down .5 if you want (not correct math but their decision where to go up/down) but do round up at some point. 4.9 should already round to 5, 4.99 / 4.999 ... even more so. :-)

Let's get other bugs like CoV amount fixed as well.

14. Originally Posted by Nibor
BTW, this is a classic example of why the userbase feels that submitting bugs through the bug reporting tool is useless. I submitted this exact bug, along with 100% recreate description, over a year ago when the new bug reporting tool was live. It was filed into a black hole for all we can tell. Someone mentions it on the forums and it gets a reply and an explanation.

Why should I bother with submitting more bug reports? I have no reason to believe anyone looks at them, and significant evidence that no one does.
Let's be even more transparent: I responded on the forums because I knew the engineer had been working on it already, because we had a bug in the system already.

Bugs go through a formal process where each one is reviewed. Forum posts have no such comprehensive review process, and in this case it was a matter of luck that I responded. If you want to make sure someone sees an issue, the reporting the bug is the better method.

15. Originally Posted by Torvaldsberg
Thank you for the explanation. I will leave the arguments about whether this the most effective way to the programmers.

Frankly, I am just glad to know that 3 plus 2 still equals 5. I have a daughter in first grade and some of the new math methods they are teaching can leave you wondering. I come from the old days where you had to memorize the basic sums and multiplication tables. Now if you ask a kid what 11 plus 14 is they start drawing pictures...
Lord help us if Turbine starts using Common Core math to code the game!

16. Originally Posted by Vargouille
Let's be even more transparent: I responded on the forums because I knew the engineer had been working on it already, because we had a bug in the system already.

Bugs go through a formal process where each one is reviewed. Forum posts have no such comprehensive review process, and in this case it was a matter of luck that I responded. If you want to make sure someone sees an issue, the reporting the bug is the better method.
Actually, the best method is to bug report it, then follow that up with a forum post about it, and to finish it off you then PM Cordovan or your other preferred dev to make sure someone is looking at your bug report. That process has never failed me.

17. Originally Posted by Vargouille
Let's be even more transparent: I responded on the forums because I knew the engineer had been working on it already, because we had a bug in the system already.

Bugs go through a formal process where each one is reviewed. Forum posts have no such comprehensive review process, and in this case it was a matter of luck that I responded. If you want to make sure someone sees an issue, the reporting the bug is the better method.
Part of the reason that you (Turbine) continue to fight the perception that forum posting is a better way to get things corrected is that once you report a bug there is no means of tracking it. It feels as if it goes into the virtual bit box. Things that might help improve that are:

1. A notification that it has been accepted into the bug reporting database.
2. A notification that is has been read/seen by an employee of Turbine.
3. A notification that it has been given to the proper department/person to investigate.
4. A notification if the bug was actually working as intended.
6. A notification that the bug is replicable and has been moved into the queue to be corrected.
7. A notification that the bug has been corrected internally.
8. A notification on what update/release/... the bug will be corrected on live.

Something along those lines... something that at least provides some feedback. It is very frustrating to submit bug after bug and not see the fruits of many of those labors. I'll admit, I don't submit a terrible amount of bugs for this very reason.

I would like to say that I do think the amount of bugs for the size of the game is actually rather small, but when some of them have been around for years without so much as a acknowledgement, it is disheartening. My personal pet peeve is the fact that you cannot complete the Orthon objective in Short Cuts and thus cannot get the 6 star favor.

18. Originally Posted by Vargouille
Let's be even more transparent: I responded on the forums because I knew the engineer had been working on it already, because we had a bug in the system already.
How long ago was he working on it? I understand your process for release takes time. (frankly tooooooo much time, not that you can do much about that.)

Because as of right now, it appears like it took well over a year for that bug to get reviewed.

19. ## Bugs and reports and bug reports

Originally Posted by Vargouille
Let's be even more transparent: I responded on the forums because I knew the engineer had been working on it already, because we had a bug in the system already.

Bugs go through a formal process where each one is reviewed. Forum posts have no such comprehensive review process, and in this case it was a matter of luck that I responded. If you want to make sure someone sees an issue, the reporting the bug is the better method.
Speaking of bugs, is there anything to report on when the in-game bug reporting will be fixed and available again? Because really, if you want people to use bug reporting then making it readily available to them at the time of finding the bugs will get you more results. Having to log out of the game or at the very least drop to desktop, open the forums and utilize the forum bug link is not so easy for some people and their computer systems as you might think. Also, you know, if you all can't or won't fix your own bug reporting tool, well, what confidence should we have that you:
B) Can fix other bugs?

20. Originally Posted by Healemup
Something along those lines... something that at least provides some feedback. It is very frustrating to submit bug after bug and not see the fruits of many of those labors. I'll admit, I don't submit a terrible amount of bugs for this very reason.
I have submitted one bug for this very reason. If they don't want to talk to me, then they likely don't want to hear from me, either. This means the time and effort I have to expend to talk to them is simply too much since they effectively ignore me from my perspective - and, let's face it, my perspective is the one that counts in this particular situation.

#### Posting Permissions

• You may not post new threads
• You may not post replies
• You may not post attachments
• You may not edit your posts
•

This form's session has expired. You need to reload the page.