[Mage] Arcane Spirit of Frost (33/0/18)

[Roxanneflowers]

This is yet another build theorycrafting of mine which relies upon the discoveries made in my (Ab)using the 5 Second Rule to maximize mana recovery from Spirit thread which you can also find elsewhere in this forum. This build isn't meant to be a bleeding edge maximal DPS at all costs build, but rather is an effort to maximize both efficiency in converting mana lost to spellcasting into damage output while at the same time reducing threat generation due to that damage output. So in that respect, rather than being a "spike" damage build designed to generate intense damage in the shortest space of time (which is desirable in PvP), this build is instead oriented around the principle of delivering steady "pressure" damage while placing reduced risk of generating threat too quickly for your designated aggro magnet to manage.

The primary spell rotation going to revolve around a Frostbolt/Arcane Missiles/repeat arrangement of 8 seconds duration for single targets, which offers a MINIMUM of 50% mana recovery from Spirit per cycle (and even more when adding in talents and Mage Armor). AoE potential for this build is delivered primarily using a fully talented Blizzard for creating a zone of "no escape" for mobs within which you can AoE them down.

Incidentally, I'm of the personal opinion that for Alliance races, Human would be a superior choice to Gnome when deliberately aiming for and equipping gear that provides a lot of Spellpower and Spirit modifiers when attempting to minimize the mana drain of spellcasting. In this case, when you're able to recover (significant quantities) of mana from Spirit while casting continuously, it is my personal opinion that a 5% boost to Spirit is preferable to a 5% boost to Intelligence ... provided that you don't want your Mage to also be an Engineer. If you do want your Mage to be an Engineer, totally go Gnome and don't look back if you're playing Alliance. For the Horde, choose either Troll or Undead to suit your preferences.



Level 60 Mage ( 33 / 0 / 18 )
https://classicdb.ch/?talent#oixEc0couZbG0Gcc

• Arcane (33 points)
  • Arcane Subtlety - Rank 2/2
    Reduces your target's resistance to all your spells by 10 and reduces the threat caused by your Arcane spells by 40%.
  • Arcane Focus - Rank 5/5
    Reduces the chance that the opponent can resist your Arcane spells by 6%.
  • Improved Arcane Missiles - Rank 5/5
    Gives you a 100% chance to avoid interruption caused by damage while channeling Arcane Missiles.
  • Magic Absorption - Rank 5/5
    Increases all resistances by 10 and causes all spells you fully resist to restore 5% of your total mana. 1 sec. cooldown.
  • Arcane Concentration - Rank 5/5
    Gives you a 10% chance of entering a Clearcasting state after any damage spell hits a target. The Clearcasting state reduces the mana cost of your next damage spell by 100%.
  • Improved Arcane Explosion - Rank 3/3
    Increases the critical strike chance of your Arcane Explosion spell by an additional 6%.
  • Arcane Meditation - Rank 3/3
    Allows 15% of your Mana regeneration to continue while casting.
  • Presence of Mind - Rank 1/1
    When activated, your next Mage spell with a casting time less than 10 sec becomes an instant cast spell.
  • Arcane Instability - Rank 3/3
    Increases your spell damage and critical strike chance by 3%.
  • Arcane Power - Rank 1/1
    When activated, your spells deal 30% more damage while costing 30% more mana to cast. This effect lasts 15 sec.
• Fire (0 points)
  • None
• Frost (18 points)
  • Frost Warding - Rank 2/2
    Increases the armor and resistances given by your Frost Armor and Ice Armor spells by 30%. In addition, gives your Frost Ward a 20% chance to reflect Frost spells and effects while active.
  • Elemental Precision - Rank 3/3
    Reduces the chance that the opponent can resist your Frost and Fire spells by 6%.
  • Ice Shards - Rank 2/5
    Increases the critical strike damage bonus of your Frost spells by 40%.
  • Permafrost - Rank 3/3
    Increases the duration of your Chill effects by 3 secs and reduces the target's speed by an additional 10%.
  • Piercing Ice - Rank 2/3
    Increases the damage done by your Frost spells by 4%.
  • Improved Blizzard - Rank 3/3
    Adds a chill effect to your Blizzard spell. This effect lowers the target's movement speed by 65%. Lasts 2 sec.
  • Frost Channeling - Rank 3/3
    Reduces the mana cost of your Frost spells by 15% and reduces the threat caused by your Frost spells by 30%.

As a leveling build, I'd recommend spending talent points like so in the build:

10-14 = Improved Arcane Missiles (for damage protection while casting)
15-17 = Elemental Precision
18-19 = Frost Warding
20-22 = Permafrost
23-24 = Ice Shards
25-27 = Improved Blizzard
28-29 = Piercing Ice
30-32 = Frost Channeling (Frost tree complete)

33-34 = Arcane Subtlety (threat reduction complete)
35-39 = Arcane Concentration
40-42 = Arcane Focus (to 3/5)
43-45 = Arcane Meditation
46-47 = Arcane Focus (to 5/5)
48-50 = Improved Arcane Explosion
51-51 = Presence of Mind
52-54 = Magic Absorption (to 3/5)
55-57 = Arcane Instability
58-58 = Arcane Power
59-60 = Magic Absorption (to 5/5)


Note that investment in Arcane Focus can be limited to only 3/5 if you prefer so as to free up 2 talent points to be invested anywhere else in the build that you might see fit … and there are a plethora of 2 talent options to choose from, so there's some reasonable flexibility there.

AoE damage production will largely come from use of Blizzard (at range) so as to create a zone of "no escape" thanks to the talents chosen in the Frost tree. Both Arcane and Frost have mana cost mitigation, threat reduction and spell resistance factors all accounted for in both trees by the time you reach Level 50 in the build plan I've provided above. Additionally, if use of (Improved) Blizzard is unwise because of a lack of mobility while channeling, you can always resort to use of Arcane Explosion, particularly after Levels 50 or 57 when the critical hit chance for (Improved) Arcane Explosion can stack up to either +6% (Level 50) or +9% (Level 57). It is important to note that Arcane Explosion CAN critically hit, while Blizzard will not (being a DoT attack).

Part of the reason for wanting to invest talent points into Frost first (for crowd control and AoE) before Arcane is so that if your Mage is also your Enchanter (which tends to happen for some strange reason) then you'll be able to solo your way into the Artisan Enchantment Trainer inside the "back door" of Uldaman rather than needing to bring a group to get past the pile of Scorpids near the trainer. This can be done with Permafrost+Improved Blizzard, but could potentially be hazardous to your health bar trying to pull off the same stunt with nothing but Mana Shield and Arcane Explosion …



While soloing against a single target, the plan is to "double tap" with Frostbolt to open before channeling Arcane Missiles. This is to prevent losing closing distance/time due to the fact that Frostbolt will only attract attention AFTER the projectile does damage, while Arcane Missiles will "instantly" draw attention when channeling begins (since Arcane Missiles places a rebuff on the target that causes them to react). The counter to this condition is to use 2x Frostbolts to open with when soloing, such that the target is very likely chilled (if not frozen) before you begin to channel Arcane Missiles, thereby increasing the amount of time it takes them to get to you before they can melee you.

I personally consider it wise to open with a "double tap" of Frostbolts just in case the first Frostbolt is Resisted, which will happen from time to time (and it's quite annoying when it does!). By sending 2x Frostbolts downrange before channeling Arcane Missiles, you reduce the possibility that your target will be able to charge at you full speed, which then in turn increases your own options for what to do ... whether that be to retreat and kite or stand and deliver (so your target can expire at your feet).



Owing to the high priority I've put on Spirit using the Frostbolt/Arcane Missiles casting rotation, I've often found it preferable to NOT use Mana Shield in a variety of situations, instead going ahead and taking the damage to health instead of to mana while leveling solo. That's because my high Spirit stat allows me to recover that lost health relatively quickly while ALSO recovering mana while in transit to my next combat engagement. With a lot of Spirit in my build, I've found myself often fully recovered by the time I've finished moving. This in turn means that using health to "soak" damage, rather than my Mana Shield spell using mana, is often the more efficient choice, except when needing to channel Blizzard or cast Frostbolt while being attacked by multiple enemies where spell pushback/interrupt becomes an issue. And in situations where the damage to health and mana starts getting heavy, I can always Eat+Drink at the same time (after combat has ended, of course) to recover all of that, just like any other Mage.

As you can probably discern from this build strategy, my own personal preference is to play as an Arcane Mage with Frost blended into the mix ... rather than as a Frost Mage with Arcane as my secondary tree. My reasoning for that is more a matter of "style" rather than one of mathematical superiority in all contexts. I'd rather ENJOY playing the my Mage the way I'd prefer to, than getting pigeon holed into playing my Mage the way I "have to" in order to join a raid group.



For anyone who is curious, my choices for investment into the Frost tree substantially match the recommendations given in the nigh invaluable In Frost We Trust video guide, albeit obviously altered and tailored to suit the needs of an Arcane build that can only invest 18-20 (or even 22 if dropping Arcane Power) talent points into Frost.

Also, when I originally began exploring the Cast+Channel spell rotation back on Darrowshire (Elysium) I was playing an Arcane/Fire build that made use of Pyroblast+Arcane Missiles for some seriously unfair damage delivery downrange (I think I astonished some teams in The Stockades). Sadly, Darrowshire "imploded" along with the rest of Elysium for me when I was just reaching the low 40 levels.

The Arcane/Fire produced a lot of damage very quickly (against targets that weren't overly resistant to fire, that is) but it suffered from a lack of crowd control to better manage engagements when things got pear shaped. This meant that the margin for error/miscalculation was punishingly small, and if you didn't react fast enough to a rapidly developing situation the result was either a big scare or a corpse run … hence why I chose to switch to an Arcane/Frost build instead to pursue the Cast+Channel spell rotation strategy. In practice, I found the "slower" tempo of Frost versus keeping Fire in the mix was a more relaxed and less stressful way to play, since Frost gives you more time to realize there's a problem and retreat with your dignity (and not a corpse yet!) intact.

[Balake]

The primary spell rotation going to revolve around a Frostbolt/Arcane Missiles/repeat arrangement of 8 seconds duration for single targets, which offers a MINIMUM of 50% mana recovery from Spirit per cycle (and even more when adding in talents and Mage Armor).

How does it reach numbers that high? IIRC, the rotation goes frost bolt (3 sec cast) into arcane missiles (5 sec channel) and repeat. There is only the 3 second window when casting frost bolt that is outside the 5 second rule, making it 3/8 uptime and not 4/8.

[Roxanneflowers]

The primary spell rotation going to revolve around a Frostbolt/Arcane Missiles/repeat arrangement of 8 seconds duration for single targets, which offers a MINIMUM of 50% mana recovery from Spirit per cycle (and even more when adding in talents and Mage Armor).

How does it reach numbers that high? IIRC, the rotation goes frost bolt (3 sec cast) into arcane missiles (5 sec channel) and repeat. There is only the 3 second window when casting frost bolt that is outside the 5 second rule, making it 3/8 uptime and not 4/8.

The answer is contained within the first thread I posted in these forums, linked HERE.

The key discovery, as enumerated in detail in the linked post, is the timing of WHEN the mana costs are paid for cast and channeled spells.
Cast spells cost mana at the END of their casting.
Channeled spells cost mana at the BEGINNING of their channeling.

if you're doing a Cast/Channel/Cast/Channel ... this is what happens for the timing of the mana costs with an untalented Frost Bolt that takes 3s to cast:

0s: Begin casting Frostbolt
3s: Cast Frostbolt, spend mana, begin 5 Second Rule debuff to mana recovery
3s: Begin channeling Arcane Missiles, spend mana, reset 5 Second Rule debuff to mana recovery
8s: Complete channeling of Arcane Missiles, 5 Second Rule of debuff to mana recovery expires
8s: Begin casting Frostbolt
11s:: Cast Frostbolt, spend mana, begin 5 Second Rule debuff to mana recovery
11s: Begin channeling Arcane Missiles, spend mana, reset 5 Second Rule debuff to mana recovery
16s: Complete channeling of Arcane Missiles, 5 Second Rule of debuff to mana recovery expires

That's the basic sequencing of events. 3 seconds then 5 seconds ... then 3 seconds then 5 seconds. The total time for the spell rotation, when repeated, is 8 seconds in duration.

Mana recovery happens every 2 seconds, but can be debuffed (up to -100%) by the 5 Second Rule of Spellcasting. So if you take a very granular view of the mana recovery ticks from spellcasting that isn't averaging things out (you only get 3 seconds out of 5 out from under the 5 Second Rule out of 8 seconds) and instead start thinking in terms of which mana recovery ticks are subject to debuffing? while doing the above rotation you get the following results when the recovery ticks are offset ever so slightly from the cast/channel times. I'm going to pick a 0.5s offset just to make this relationship clear(er):

0.5s: 100%
2.5s: 100%
4.5s: debuffed
6.5s: debuffed
8.5s: 100%
10.5s: 100%
12.5s: debuffed
14.5s: debuffed
16.5s = return to start @ 0.5s and repeat as needed



Here's what that looks like when you put the two tables together into one for easier viewing and comprehension of WHAT is happening WHEN in this sequencing:

0s: Begin casting Frostbolt

• 0.5s: 100% mana recovery from Spirit
• 2.5s: 100% mana recovery from Spirit

3s: Cast Frostbolt, spend mana, begin 5 Second Rule debuff to mana recovery
3s: Begin channeling Arcane Missiles, spend mana, reset 5 Second Rule debuff to mana recovery

• 4.5s: mana recovery from Spirit debuffed
• 6.5s: mana recovery from Spirit debuffed

8s: Complete channeling of Arcane Missiles, 5 Second Rule of debuff to mana recovery expires
8s: Begin casting Frostbolt

• 8.5s: 100% mana recovery from Spirit
• 10.5s: 100% mana recovery from Spirit

11s:: Cast Frostbolt, spend mana, begin 5 Second Rule debuff to mana recovery
11s: Begin channeling Arcane Missiles, spend mana, reset 5 Second Rule debuff to mana recovery

• 12.5s: mana recovery from Spirit debuffed
• 14.5s: mana recovery from Spirit debuffed

16s: Complete channeling of Arcane Missiles, 5 Second Rule of debuff to mana recovery expires

• ... repeat ...

In other words, although the 5 Second Rule of Spellcasting has a debuff that lasts for 5 seconds (as advertised) ... mana recovery "ticks" every 2 seconds. This means that although the 5 Second Rule of Spellcasting lasts for 5 seconds (as advertised), it's only going to affect up to 2 ticks of mana recovery (2s, 4s) and not 3 ticks (2s, 4s, 6s). This is then why with an 8s overall Cast+Channel spell rotation you're really only going to have 2 mana recovery ticks getting debuffed per rotation, thanks to the timing of exactly WHEN the mana recovery ticks occur.

If mana recovery from Spirit were a continuous phenomenon, rather than a granular "quantified" recovery via regularly scheduled ticks (the schedule does not change), then the interpretation of 3s every 8s out from under the 5 Second Rule of Spellcasting would be the correct one ... but that's not the way it works game mechanically under the hood. Instead, there's an inter(re)action between the inexorable mana recovery every 2s ticks and whether the 5 Second Rule of Spellcasting debuff is in effect WHEN those 2s ticks occur. Since the gap between the mana recovery ticks is 2 seconds, and the 5 Second Rule of Spellcasting lasts for 5 seconds, the way it works in practice is that spending mana will prevent mana recovery for the next 2 (not 3! only 2) mana recovery ticks ... provided that you aren't spending mana again 5 seconds later after you last spent mana.

So if the delay between spending mana is 8s because of the "shape" of when mana gets spent by your spell rotation (Cast spells costs at the end, Channeled spells cost at the beginning, putting the mana costs effectively back to back when doing Cast+Channel on repeat) only 2 mana recovery ticks out of 4 get debuffed by the 5 Second Rule of Spellcasting, thanks to the sheer TIMING of when mana costs for spells happen during the 8s spell rotation.

This is why I call this particular insight a way to (Ab)Use the 5 Second Rule of Spellcasting to your advantage for being able to recover mana from Spirit while continuously casting the entire time. HOW it is able to do so is not exactly intuitive to grasp (thanks to the 3s out of 5s "logical trap" that you cite) since it requires going that extra step to account for WHEN mana recovery ticks happen so as to determine WHICH mana recovery ticks get debuffed WHEN by spending mana during the spell rotation ... and it's precisely that manner of "clever shaping of mana cost via timing" that allows the performance I've described to be possible in the way that I've described. It's one of those things that once explained and laid out becomes obvious, but until then it's hidden and obscured behind the generalized assumptions made for spellcasting using Cast spells only.

So I honestly recommend that you read what I wrote about Mana recovery from Spirit while continuously casting... because I've tested it AND IT WORKS(!). Like you, I'd started from a default understanding that 5 Seconds Means 5 Seconds of lockout (which it does, technically) ... but there's another layer of interaction with the timing of the 2s between mana recovery ticks after spending mana ... and 5 Seconds does not equal THREE mana recovery ticks, it only debuffs TWO ... if you don't spend mana faster than every 6 seconds(!). And with a 3+5=8s long spell rotation of Cast+Channel, you'll only be losing 2 mana recovery ticks out of 4 to debuffing every 8s ... not 2.5 mana recovery ticks out of 4 to debuffing every 8s.



I hope that helps with understanding the insight that underpins the planning behind the spell rotation I plan to use for this build and WHY it works the way it does. I'd even go so far as to suspect that the original Blizzard Devs might have expected Players to discover this inter(re)action on their own and conclude that using Arcane Missiles as part of a repeating spell rotation that allowed for mana recovery from Spirit while continuously casting was net mana cost CHEAPER than just doing what actually wound up happening with the One Button Mashing Spam™ of simply casting the same spell over and over again until OoM, never once getting out from under the 5 Second Rule of Spellcasting the entire time.

Unfortunately, Players tend to be a lazy bunch who dogpile into doing the simplest thing that requires the least investment of time, effort, analysis (and even strategy) that can be repeated endlessly while requiring the least amount of thought/effort to accomplish their goal(s). We tend to be averse to complexities and angle for the Path Of Least Resistance to our rewards as much as possible ... and once the Conventional Wisdom™ around a particular set of assumptions or notions on How To Play harden, they tend to wind up getting set in concrete (see: You're Doing It Wrong admonishments) ... until someone comes along to challenge the underlying set of assumptions and discovers that there's more to what's going on (game mechanically, under the hood) than what had been commonly assumed or believed.



But first, that discovery has to be made.
Then it has to be tested.
Then it has to be shared.
And then the REAL challenge begins ... it has to be believed(!) before it can be ... accepted.



I've tested this discovery on two private servers (Elysium and Symmetry) and it worked astonishingly well, in the way I've described ... so much so that I wound up putting a lot of value in +Spirit on my (Night Elf) Priest, Mage (as shown here) and Warlock builds (yes, even Warlock!) because of how much of a positive effect it had to use a Cast+Channel spell rotation with each of those classes to stave off mana starvation and going OoM. Being able to stay "flush with resources" even after heavy combat was simply game changing for me and how I was able (and therefore, willing) to play my own characters.

However, you don't have to take my word for it on pure faith. You can test the proposition yourself, without needing to spend on a respec, if you've got a character who can do a Cast+Channel repeating rotation. There are Target Dummies in capital cities that you can easily test the rotation strategy on and simply time yourself on how quickly you go OoM using different spell rotations and no changes in talents. I'm pretty sure you'll be as surprised at the difference in stopwatch endurance while continuously casting as I was when I started doing the Cast+Channel rotation while hunting out in the open world. You'll also be able to see for yourself just HOW the mana recovery from Spirit system inter(re)acts with the 5 Second Rule of Spellcasting and whether what I'm saying here has any merit to it or not. I figure that might be worth ... oh, say ... 5-10 minutes of your time to test out for yourself if you have any doubts.

[Balake]

Ah I see, I thought mana worked with smooth regen that was only represented visually by ticks, and not actual ticks every 2 seconds like energy. so it requires counting the regen ticks just like a feral druid would do when powershifting. If you start the rotation 0.5 seconds before the regen tick, you indeed get 50% uptime outside the mp5 rule which is great. But if you screw up and start it let's say 1.5 seconds before the regen tick, this happens:

0s: Begin casting Frostbolt
1.5s: 100% mana recovery from Spirit
3s: Cast Frostbolt, spend mana, begin 5 Second Rule debuff to mana recovery
3s: Begin channeling Arcane Missiles, spend mana, reset 5 Second Rule debuff to mana recovery
3.5s: mana recovery from Spirit debuffed
5.5s: mana recovery from Spirit debuffed
7.5s: mana recovery from Spirit debuffed
8s: Complete channeling of Arcane Missiles, 5 Second Rule of debuff to mana recovery expires
8s: Begin casting Frostbolt
9.5s: 100% mana recovery from Spirit
11s:: Cast Frostbolt, spend mana, begin 5 Second Rule debuff to mana recovery
11s: Begin channeling Arcane Missiles, spend mana, reset 5 Second Rule debuff to mana recovery
11.5s: mana recovery from Spirit debuffed
13.5s: mana recovery from Spirit debuffed
15.5s: mana recovery from Spirit Debuffed
16s: Complete channeling of Arcane Missiles, 5 Second Rule of debuff to mana recovery expires
... repeat ...

Messing up the timings will lower the regen tick uptime to 25%. To do this strategy effectively, one needs a 2 second tick tracker to make sure they don't get badly time it and drop their mana regen from 50% to 25%.

Now, if one was freely casting like in a dungeon group with no mobs attacking them. They can simply time it right the first time, and they will stay at 50% for the whole fight because the rotation lasts an even number (8 seconds) which translates to 4 ticks. But if for example they time it right, but they get hit by something and arcane missiles gets pushed back to 4 seconds the rotation becomes 7 seconds long, and they need to wait another 1 second to start the rotation again or they will only get 25% tick uptime from the next rotation.

[Roxanneflowers]

Ah I see, I thought mana worked with smooth regen that was only represented visually by ticks, and not actual ticks every 2 seconds like energy. so it requires counting the regen ticks just like a feral druid would do when powershifting.

Same principles apply, yes, because it's a common system for all spellcasters.

If you start the rotation 0.5 seconds before the regen tick, you indeed get 50% uptime outside the mp5 rule which is great. But if you screw up and start it let's say 1.5 seconds before the regen tick, this happens:

If you get the timing EXACTLY WRONG there is an edge case possibility that you might miss a mana recovery tick and only get 1 out of 4 instead of 2 out of 4 ... but in terms of actual gameplay experience, the odds of that happening with an 8s spell rotation is actually pretty darn miniscule (you have to be absurdly unlucky for it to happen, and it's extremely difficult to sustain that unluck for duration).

Of course, if you throw in 5/5 Improved Frostbolt so as to shorten the casting time of Frostbolt from 3s down to 2.5s, causing the overall spell rotation to drop from 8s to 7.5s, the opportunity for that kind of "miss a recovery tick" condition increases because you're not synchronizing the duration of your spell rotation to the 2s mana recovery tick schedule.

To do this strategy effectively, one needs a 2 second tick tracker to make sure they don't get badly time it and drop their mana regen from 50% to 25%.

Not in my experience, mainly because people don't cast exclusively on the 0.0/0.5/1.0/1.5s timing relative to the mana recovery ticks. People cast at any old time they feel like it and then just keep casting, and lag factors (network, server, etc.) can always introduce additional variables to the precision that cannot be accounted for by Player UI control. So while theoretically possible, it just simply winds up being something that rarely (if ever) happens, and even when (if?) it does the condition doesn't persist all that long, making the condition you're talking about relatively unique in the sense that it's absurdly hard to sustain over multiple rotations.

In other words ... it's really not worth worrying about that much, because it can only happen with an EXACTLY wrong synchronization that is unlikely to be sustained in actual gameplay because UI reaction speeds from user inputs have tolerances of reaction times. You can train yourself as a Player to get "better" at your reaction times, but it's unlikely you'll be "perfect" consistently forever. So any "statistical noise" in the reaction speed(s) for chaining the spells will knock you off that EXACTLY WRONG conditional scenario you've described to prevent it from recurring repeatedly.

So yes, it CAN (theoretically) happen ... but if it does it's a One Off condition of such a vanishingly small likelihood that it would be an aberration for it to happen if it ever does because of the variables involved in actual networked gameplay. That's why I minimize the possibility the way that I am here as a factor to worry about.

[Balake]

No, it is not a small chance. When you start the rotation, there is a 50% chance that you will get 1 tick while casting frost bolt and 3 ticks while casting arcane missiles (for 25% mana regen) and a 50% chance of 2 ticks in frost bolt 2 ticks in arcane missiles (For 50% mana regen).
t=0 sec we get a mana regen tick. We have until t=2 sec for the next mana regen tick. If you start the rotation between t=0 and t=1, you get 50% mana regen uptime (the ideal situation). If you start the rotation between t=1 and t=2, you get 25% mana regen uptime (the bad situation). If someone is just casting it randomly without tracking mana ticks, they'll get it right half the time which is not ideal.
So getting 50% mana regen uptime is actually not the bare minimum you can get that can be increased through talents and mage armor. It's actually what you can get if you correctly time your mana regen ticks. If you do it wrong every single time, you get 25% mana regen uptime. But they average out to exactly 37.5% or 3/8 which is what it would have been if mana regen was counted smoothly instead of in ticks in the first place surprisingly to me.

[Roxanneflowers]

No, it is not a small chance.

I played 3 different classes into the low 40s on Darrowshire using this method. What you describe was NOT common AT ALL in my experience, let alone as common as you're trying to make it seem here.

But ... in order to prove that, you'd need to test it, wouldn't you? There are target dummies in capital cities. Test it. You don't have to take my word for it ... test it. See for yourself how right/wrong you are in your assertions that run counter to my 20+ levels of experience on 3 different classes using this method. It's not that difficult a test to do.

Simply do your normal (for you) spell rotation and time how long it takes you to go OoM.
Then switch over to doing a Cast+Channel rotation and time how long it takes you to go OoM.

If the latter is longer than the former ... well ... you ought to know why that might be by now.

Test it. See for yourself.
Even better yet, if you fraps record your testing you can play it back and count the mana ticks as they happen in actual gameplay (as opposed to merely theory). You'll be able to see directly how reliably you can get only 1 mana recovery tick per rotation as opposed to 2.

[Balake]

There is no denying that it's going to take much longer to go oom with this strategy, and also do more damage before doing so. I don't need to test out to know that, but it's like comparing apples to oranges because each of these is better for different things. For a mage, yes this would lead to more damage being dealt before the mana bar empties, but it leads to longer fights and more damage taken and requires different itemization than what mages normally aim for.

However, that is not what I have an issue with.This build is good for the role that it was meant to fill, and that's rotation that will let you keep casting for as long as possible without ooming. I'm concerned about how much effort this actually takes to pull off. Regarding this specific issue, the theory is no different from practice. As I showed earlier, there is a 1 sec out of 2 seconds window each mana tick to start the rotation. If you start it during that window you get 50% mana tick uptime, but if you start it during the other 1 sec window of the 2 seconds tick, you will only get 25% mana tick uptime. If done randomly, the math says the player will get it wrong 50% of the time. I tried it by starting with a full mana bar, casting shadow bolt then drain life then a shadow bolt, and looking at how much mana ticks I get during this last shadow bolt. I noticed close results. I sometimes got 1 tick and other times 2 ticks and they seemed equally likely and not a rare occurance.

[Roxanneflowers]

As I showed earlier, there is a 1 sec out of 2 seconds window each mana tick to start the rotation. If you start it during that window you get 50% mana tick uptime, but if you start it during the other 1 sec window of the 2 seconds tick, you will only get 25% mana tick uptime. If done randomly, the math says the player will get it wrong 50% of the time. I tried it by starting with a full mana bar, casting shadow bolt then drain life then a shadow bolt, and looking at how much mana ticks I get during this last shadow bolt. I noticed close results. I sometimes got 1 tick and other times 2 ticks and they seemed equally likely and not a rare occurrence.

That was not my experience with it on either Darrowshire nor Symmetry, so I wonder if there's some other unaccounted for factor involved.