Wizard.
--
Derek
Game info and change log: http://sporkhack.com
Beta Server: telnet://sporkhack.com
IRC: irc.freenode.net, #sporkhack
That probably depends on the player, but I'm with David Goldfarb; if you
put a gun against my head and told me to ascend or die with my character,
I'd like a dwarven valkyrie, please.
--
David Damerell <dame...@chiark.greenend.org.uk> flcl?
Today is Saturday, March - a weekend.
The class most suited to their knowledge, skills, and style of play.
Can't see why. Wizards start with MR and get Magicbane in Minetown*;
that's a game win right there with those two (no fear of polytraps, wands of
striking, magic missile, or death, plus infinite one-turn Elbereth).
They can wear all the armor a Valkyrie can, plus cast every spell at 0% and
reliably write any scroll they need.
Their quest nemesis is a howling wuss, there's only the one drawbridge in
the game to worry about, and the quest artifact is completely broken.
Elven wizards start with an instrument, and as long as it's tonal can
speed-dive to the Castle (note: NOT DfV) and trivially obtain the wand while
guarded by the guaranteed earth scrolls from Sokoban.
About the only concern they have is getting poison resistance, and if
you're sufficiently worried about that, well, orcs are available.
* If someone doesn't know how to convert hostile attended altars, their risk of
death is too high for choice of class to matter.
Adman
That's a bit of a non-answer. Personal preference plays a part, but
there's no denying that classes have inherent merits. Who'd rather be a
Caveman than a Valkyrie?
Well, our player is "spoiled", but presumably no Marvin - otherwise the
difference between ascension rates with classes are too small to pick out.
I am pretty well spoiled but no Marvin, so I imagine them as being a bit
like me.
So it seems to me what will carry off most of their characters is infant
mortality; deaths in the opening game. MR isn't doing a lot there, and by
the time they get to Minetown they're already over much of the opening
game hump.
I'd rather have the +3 small shield, the bucket of hitpoints, the good
melee (and ability to practice daggers from the get-go), starting food
(conserve prayer or generate pets), ease of getting Excalibur, etc. Sure,
the Wizard will be stronger later on; but who cares about that? Later on
the game is likely in the bag anyway.
The question is really unanswerable in any more detail. It's all
preference. Even "easiest" is undefined, here. Easiest for whom? To
achieve what goal?
For an answer devoid of personal preference - Download
the ascensions from NOA and build a table. Pick the
class with the the highest ascension percentage.
For an answer based on popularity - Download the
ascensions from NOA and build a table. Pick the class
with the highest number of ascensions.
For an answer based on your own playing record -
Process a record of your games. Pick the class you
have the most ascensions in (popularity in your own
view) or the highest percentage. Lacking ascensions
pick the class with the most entries on your high score
list.
There are folks who prefer spell casters and folks who
prefer combat wombats. Their choices as posted tend
to be wizard or valkyrie.
I'm currently playing more knights because of the
weaker missile skills so I need higher reliance on melee
in the early game before they become magic missile
blasters in the late game. I'm struggling to survive with
that much melee because I've previously learned to use
missiles and spells.
> For a player who is spoiled, which class is easiest?
Foodless, atheist tourist.
>> For a player who is spoiled, which class is easiest?
> Foodless, atheist tourist.
You are an evil, evil human being, Haakon Studebaker! ;-)
You did make me laugh, though.
xanthian.
At least you wrote that answer to someone who will take
it in the way it was intended.
Another answer might be "any class, as long as you
include twelve conducts".
> You are an evil, evil human being, Haakon Studebaker! ;-)
>
> You did make me laugh, though.
Well he DID say spoiled.
When you look at the numbers on NAO, valks and barbs have the highest
ascension rates for both new players and experienced ones.
Wizards have the highest disparity; they're among the lowest overall,
but for experienced players they're in the top 5 and not too far
behind the big ones.
But basically, valks and barbs are easiest by the numbers. For
individuals, that may vary.
Indeed. For someone with complete access to spoilers, I'd still give the
slight edge to Valk, though. The Barb's biggest starting advantage for
an unspoiled player with little playing experience -- XL 1 poison
resistance -- is of less account if you know what you can eat. (However,
it does prevent instadeaths from poisonous spiked pits and poisoned
missiles. And an early potion of sickness won't ruin your day.)
I'd say the three words that put Valk over the top are Orb of Weight,
though. Half physical and half spell damage in one (bulky) package is
just too good a deal. My first 3.1.3 ascension was a Valk, and my first
3.4.3 ascension also would've been one if I'd paid a smidge more
attention when one of Rodney's pet krakens got friendly with me.
It's a good deal, but really - once you've *got* the Orb, you should have
the game in the bag with or without it.
--
David Damerell <dame...@chiark.greenend.org.uk> Oil is for sissies
Today is Sunday, March - a weekend.
I thought that Excalibur and Expert Dagger was what puts the Valk over
the Barbarian? Valks start with a long sword, and can be lawful, so
they can easily get Excalibur in the early game. They also can get
Expert Dagger, which is important because Barbarians don't really get
any decent ranged weapon options.
For experienced, spoiled players, quest artifacts are essentially
irrelevant to ascension percentage. If you've made it through the
quest, you're going to ascend (barring tragic misplay).
Not accurate info. I've gone over this here before (around 2006 or so),
but this assumes that every game is being played to win, which isn't
true on NAO due to conducts, etc.
A quick perlscript across the latest logfile from NAO produces the
following, however:
Role: Cav Ascs: 14
Role: Kni Ascs: 26
Role: Ran Ascs: 42
Role: Hea Ascs: 47
Role: Rog Ascs: 50
Role: Pri Ascs: 60
Role: Mon Ascs: 61
Role: Arc Ascs: 69
Role: Tou Ascs: 87
Role: Sam Ascs: 95
Role: Bar Ascs: 104
Role: Val Ascs: 426
Role: Wiz Ascs: 437
The script checks the first ascension on record for each player on
version 3.4.3, and totals up which role it was. Since ascenders on
NAO can be assumed to be "spoiled" to some degree, this is significantly
more accurate than straightforward ascension percentage.
Interestingly, last time I did this, Valk was about halfway inbetween
Barb and Wiz; but it's crept up a lot since then. It's rather safe to
assume that the "right" answer, though, is either Valkyrie _or_ Wizard,
and as to exactly why? Refer to either my post detailing the advantages
of the Wizard (who has so many tools at their disposal that it's hard to
die in the first place), or other peoples' posts detailing the Valkyrie
(who has a big advantage in the early game).
The former is important, the latter less so. Both Barbs and Valks should
be juggernauts, walking all over everything except Demo, without
bothering to use ranged weapons at all.
My first ascensions were with Barbs, but that was back when dipping for
Excalibur was broken for anything except Knights. Now, I'd start on
Valks.
Richard
Significantly _less_, IMO, since there's a massive selection bias
here. Wizards are much more fun to play than other roles, so you get
more people playing them.
You and I and a bunch of others sliced the NAO numbers a bazillion
ways in an earlier thread. Given that, I'm pretty confident that for
the vast majority of (spoiled or not) players, Val, Bar, Sam (and Cav,
Kni to a lesser degree, and Ran as well for experienced players) are
the easiest in terms of having the highest ascension % (new or
experienced doesn't matter much), and that Wiz is in the bottom half
ease-wise even for very experienced players.
In fact, for players with 20+ ascensions, wizards had the _lowest_ asc
% of any role. That's a small group of players so not a high degree
of confidence, but the group of players with 10-20 ascensions is much
larger and Wizards still ranked 3rd-hardest (with Arc and Mon harder).
I even sliced it up by "best role" to see if there was a measurable
play-style difference or not. Whether a player has their best asc% as
a Val, Bar, Wiz, Hea, Arc, or Tou didn't seem to matter much: their
next-best roles were pretty much Val, Bar, Sam (or occasionally one of
those would slip to 4th or 5th with one of the aforementioned Cav/Kni/
Ran in play).
Here's one of the posts I made in that "Role difficulty and number of
ascensions" thread:
Okay, numbers with all quits/escapes for max DL=1 eliminated, again
only for players with at least 3 ascensions.
Anything else anyone wants?
Ascension percentage for players....
Up to (and including) 1st ascension:
2.82 Val
2.48 Bar
2.31 Sam
1.64 Tou
1.61 Arc
1.36 Kni
1.18 Cav
1.17 Rog
1.09 Wiz
0.93 Mon
0.87 Pri
0.79 Ran
0.51 Hea
2nd ascension (too high variance to show much other than that after
the 1st ascension, rates _skyrocket_):
19.29 Val
13.30 Bar
8.51 Sam
6.07 Kni
4.27 Mon
4.19 Tou
3.97 Wiz
3.73 Cav
3.06 Pri
2.87 Rog
2.68 Hea
2.62 Ran
2.09 Arc
From 3rd to 10th ascension:
18.22 Val
16.83 Bar
13.69 Sam
11.18 Kni
10.38 Cav
7.39 Ran
6.29 Hea
6.16 Mon
6.13 Rog
6.03 Pri
5.96 Tou
5.19 Wiz
3.33 Arc
For 10th-20th ascensions:
23.78 Sam
21.03 Val
18.12 Cav
17.11 Bar
16.90 Kni
14.29 Tou
13.46 Ran
13.12 Pri
11.95 Hea
10.40 Rog
8.04 Wiz
7.97 Arc
7.74 Mon
For 20+th ascensions:
36.36 Ran
33.19 Val
33.16 Kni
31.94 Bar
31.47 Cav
30.14 Mon
24.90 Rog
23.57 Hea
22.01 Pri
21.71 Tou
20.99 Sam
19.29 Arc
12.39 Wiz
Top 4 other roles by asc% for players whose best asc% is for the first
role:
Arc Sam 28.92% Val 25.15% Bar 24.14% Kni 18.75%
Bar Cav 16.89% Kni 16.75% Val 16.49% Sam 16.01%
Cav Val 35.50% Sam 32.07% Ran 27.59% Bar 26.67%
Hea Kni 37.76% Ran 36.80% Cav 35.78% Val 32.64%
Kni Val 16.47% Bar 13.46% Cav 12.46% Sam 11.02%
Mon Val 27.94% Bar 24.49% Sam 21.43% Ran 15.67%
Pri Bar 42.19% Cav 35.19% Val 34.29% Sam 22.79%
Ran Cav 30.30% Bar 22.84% Val 21.13% Sam 20.97%
Rog Val 25.23% Bar 25.00% Sam 16.30% Cav 14.95%
Sam Val 23.76% Bar 20.94% Kni 13.84% Ran 13.73%
Tou Val 25.71% Bar 24.11% Sam 19.18% Ran 13.74%
Val Cav 11.72% Kni 11.35% Bar 11.12% Sam 9.97%
Wiz Cav 25.68% Val 20.42% Bar 20.00% Sam 13.11%
Full table of all 13 (wrap-unfriendly):
Arc Sam 28.92% Val 25.15% Bar 24.14% Kni 18.75% Cav 16.00% Mon 11.19%
Hea 11.04% Wiz 9.23% Tou 7.75% Pri 7.56% Rog 7.25% Ran 6.90%
Bar Cav 16.89% Kni 16.75% Val 16.49% Sam 16.01% Ran 11.33% Mon 10.78%
Tou 9.32% Arc 9.04% Rog 8.67% Wiz 8.00% Hea 7.89% Pri 7.03%
Cav Val 35.50% Sam 32.07% Ran 27.59% Bar 26.67% Kni 21.17% Tou 17.90%
Hea 15.67% Mon 11.99% Wiz 11.85% Arc 11.43% Pri 11.37% Rog 11.26%
Hea Kni 37.76% Ran 36.80% Cav 35.78% Val 32.64% Sam 32.43% Bar 30.91%
Rog 28.47% Pri 23.15% Wiz 19.12% Tou 17.89% Mon 17.11% Arc 15.85%
Kni Val 16.47% Bar 13.46% Cav 12.46% Sam 11.02% Hea 7.75% Ran 7.72%
Tou 7.71% Pri 6.89% Arc 6.73% Wiz 5.96% Rog 5.67% Mon 4.86%
Mon Val 27.94% Bar 24.49% Sam 21.43% Ran 15.67% Cav 15.00% Kni 12.50%
Rog 9.55% Wiz 9.03% Tou 8.86% Pri 8.25% Hea 6.93% Arc 6.82%
Pri Bar 42.19% Cav 35.19% Val 34.29% Sam 22.79% Ran 21.65% Kni 21.57%
Mon 16.55% Hea 15.95% Rog 14.14% Wiz 13.40% Tou 12.50% Arc 8.90%
Ran Cav 30.30% Bar 22.84% Val 21.13% Sam 20.97% Hea 16.90% Kni 16.13%
Rog 13.73% Pri 13.73% Mon 13.67% Arc 10.92% Tou 9.84% Wiz 7.26%
Rog Val 25.23% Bar 25.00% Sam 16.30% Cav 14.95% Ran 10.34% Kni 9.82%
Hea 9.50% Mon 8.38% Pri 7.59% Wiz 7.29% Arc 4.91% Tou 4.50%
Sam Val 23.76% Bar 20.94% Kni 13.84% Ran 13.73% Cav 13.02% Rog 11.26%
Pri 9.34% Tou 9.30% Arc 9.23% Hea 8.71% Wiz 7.27% Mon 7.18%
Tou Val 25.71% Bar 24.11% Sam 19.18% Ran 13.74% Hea 10.81% Cav 10.27%
Kni 9.54% Arc 9.18% Rog 7.72% Mon 6.34% Pri 5.84% Wiz 5.44%
Val Cav 11.72% Kni 11.35% Bar 11.12% Sam 9.97% Ran 6.98% Pri 6.59% Hea
6.29% Tou 5.93% Rog 5.89% Mon 5.68% Wiz 3.73% Arc 2.51%
Wiz Cav 25.68% Val 20.42% Bar 20.00% Sam 13.11% Rog 11.54% Tou 9.33%
Kni 8.94% Pri 7.42% Ran 7.36% Hea 6.98% Mon 6.57% Arc 6.24%
That selection bias exists for any numbers you'd care to pull off NAO;
it's the reason Cavemen and Knights show up so haphazardly in all
statistics, because they just aren't played enough.
Perhaps you may wish to consider that more people play wizards because
they're more fun ... because they're easier?
> You and I and a bunch of others sliced the NAO numbers a bazillion
> ways in an earlier thread. Given that, I'm pretty confident
And I'm pretty confident of my own conclusions too. I bailed out on
that thread finally because I just don't have the time to hash over
statistics again -- but if you dig back far enough there is a big ol'
spreadsheet that was posted onto akrasiac.org that spent a great deal of
time avoiding ascension percentage (though it listed them as well).
> In fact, for players with 20+ ascensions, wizards had the _lowest_ asc
> % of any role.
That is because people with 20+ ascensions start doing Stupid Nethack
Tricks, and one of them was/is Dig For Victory, which has a high death
rate but needs the Wizard as its ideal class. In fact, for awhile
someone was running something called 'scorefest', which was the attempt
to die or escape with the most points within a half hour. That killed
so many wizards it wasn't even funny, and had people with huge ascension
numbers in it as well.
Ascension percentage is the worst possible method to measure anything
on NAO, because there's just no way to separate out the people
trying to screw around from the people trying to win. That's why I
bothered to dig through and get "first ascension per player", which is a
number independent of # of attempts and, by definition, independent of # of
deaths from screwing around.
Statistics are good as far as they go, but common sense has to be
applied as well, and when you get numbers that say "for very good
players, wizards are hard", that flies directly in the face of common
sense and indicates that there is a factor you haven't allowed for -- in
this case, even after filtering out quits and escapes, Wizards are the
'screwaround' class, generating enough "not-trying-to-win" deaths to render
ascension percentage meaningless as a metric no matter how you slice up
the numbers.
I blame Eidolos.
> > Significantly _less_, IMO, since there's a massive selection bias
> > here. Wizards are much more fun to play than other roles, so you get
> > more people playing them.
>
> That selection bias exists for any numbers you'd care to pull off NAO;
> it's the reason Cavemen and Knights show up so haphazardly in all
> statistics, because they just aren't played enough.
Right. That makes counting stats useless. You need to use some sort
of rate stat to have anything remotely meaningful. This is pretty
basic statistics here--if 20 times as many valks and wiz as cav or kni
are started, you can't just add up totals and get anything meaningful.
As long as you're looking at totals, your numbers are completely
worthless. Wizards account for a huge number of games on NAO, so
they're going to amass a lot of everything. It's like looking at how
many career TDs a 15-year veteran QB has vs. a 3rd year player and
saying "hey, look, he's scored more he must be better!"
There isn't any significant cross-section of players that succeeds
with wizards at anything remotely approaching the level of success
they have with the big combat-wombats (Val/Bar/Sam).
You can look at very experienced players. You can look at middling
players (say, those with <12 ascensions) who are more likely to still
be enamored of ascension and not yet chasing DfV and such. You can
look at players who should know they're in the middle of an ascension
streak (previous 2+ games were ascensions), and thus presumably much
less likely to mess around. Wizards just don't survive nearly as much
as the big wombats.
Intuitively, that makes sense--things like starting with MR are great
helps, but not in the very early game where death is most likely. By
10+ years of general wisdom on this group, it makes sense--Valks and
Barbs always come up most often in the discussions of the easiest
roles. By my own experience it makes sense.
> > You and I and a bunch of others sliced the NAO numbers a bazillion
> > ways in an earlier thread. Given that, I'm pretty confident
>
> And I'm pretty confident of my own conclusions too. I bailed out on
> that thread finally because I just don't have the time to hash over
> statistics again -- but if you dig back far enough there is a big ol'
> spreadsheet that was posted onto akrasiac.org that spent a great deal of
> time avoiding ascension percentage (though it listed them as well).
Right. Then as now I don't understand the reason that you would go
through such contortions to avoid using a stat that seems to be a lot
more relevant to what you're studying, aside from the obvious "I don't
like what it tells me so I'm going to futz with the numbers".
> > In fact, for players with 20+ ascensions, wizards had the _lowest_ asc
> > % of any role.
>
> That is because people with 20+ ascensions start doing Stupid Nethack
> Tricks, and one of them was/is Dig For Victory, which has a high death
> rate but needs the Wizard as its ideal class.
But as I pointed out, the same is true for people with 10-20
ascensions.
I also looked at only players with <12 ascensions, and threw out the
learning curve/screwing around newbie stuff (pre-3rd ascension). Val/
Bar/Sam wound up easiest, with Mon/Wiz/Arc at the bottom.
Wizard's don't fare well with any level of player. Is _everyone_
diving for victory?
I also looked at only people in the middle of ascension streaks
(previous 2+ games were ascensions) and got similar results. Are
people screwing around mid-streak as well??
At some point don't you have to look to Occam's Razor and decide that
the reason nobody has great success with wizards isn't because there's
some different weird explanation for each group, but that wizards
simply aren't as survivable early on as roles with more AC and HP?
> In fact, for awhile
> someone was running something called 'scorefest', which was the attempt
> to die or escape with the most points within a half hour. That killed
> so many wizards it wasn't even funny, and had people with huge ascension
> numbers in it as well.
The only scorefest game I can find (looking up the ones listed at
http://nethack.wikia.com/wiki/Nethack.alt.org_weekly_scorefest ) is
for a valk. Before I even considered whether this was meaningful, I'd
want to know how many games it accounted for relative to the total and
what the distribution of various roles was in these games.
> Ascension percentage is the worst possible method to measure anything
> on NAO, because there's just no way to separate out the people
> trying to screw around from the people trying to win. That's why I
> bothered to dig through and get "first ascension per player", which is a
> number independent of # of attempts and, by definition, independent of # of
> deaths from screwing around.
But it's a rate stat. It's useless. My "asc% for only the first
ascension" is _precisely_ as effective at using only the first
ascension (less likely to be screwy) but _also_ controls for the
number of games started with each role. And it says exactly the same
thing as all the other methods: Valk smash (or the moral equivalent)
is the way to go.
> Statistics are good as far as they go, but common sense has to be
> applied as well, and when you get numbers that say "for very good
> players, wizards are hard", that flies directly in the face of common
> sense
I'm not sure if this is wishful thinking or polemics. The common
sense belief is that the crunchy combat roles are much easier. The
numbers support that pretty obviously. Don't conflate your personal
foibles with common sense.
[...]
> By 10+ years of general wisdom on this group, it makes
> sense--Valks and Barbs always come up most often in the
> discussions of the easiest roles. By my own experience it
> makes sense.
Just curious, but why Valks and Barbs, rather than Valks and
Samurai? All three are powerful in melee, Valks and Samurai
have the advantage of being able to effectively use a ranged
weapon as well, and early speed, and the Samurai starts with
pretty good armor as well. About the only disadvantage of a
Samurai is that its quest nemesis has the Tsurugi of Muramasa,
which can result in an instant death if you aren't careful.
And how big a role does the quest nemesis play in ascensions?
Valks and Wizards certainly have an easy one---could that also
be an argument in their favor.
--
James Kanze (GABI Software) email:james...@gmail.com
Conseils en informatique orientée objet/
Beratung in objektorientierter Datenverarbeitung
9 place Sémard, 78210 St.-Cyr-l'École, France, +33 (0)1 30 23 00 34
Second question first; I don't think it plays a major part. A spoiled
player would know that the Samurai quest artifact isn't particularly
useful and will happily put it off until after they have Orcus' wand
of death and Vlad's amulet of life saving.
For the first question, I think both Valkyries and Barbarians have
small advantages over Samurai in the early game. Barbs get their
poison resistance, Valks get easy Excalibur, and both get slightly
better HP growth than a Samurai. They're not big differences, and in
the end a decent player will ascend close enough to 100% with any of
these classes, but I'd say that Samurai are slightly tougher overall.
All three are top (analyzing the NAO data). But the former three are
mostly discussed and taken as example in RGRN, which is what sjdevnull
seems to have said.
> And how big a role does the quest nemesis play in ascensions?
> Valks and Wizards certainly have an easy one---could that also
> be an argument in their favor.
My personal opinion is that the quest nemesis difficulty should not be
a problem, anyway. Quite all nemeses have their weaknesses, and if you
know about those you'll have all the time you like to get what you need
before you decide to meet them.
If, OTOH, one doesn't know about a nemesis' strength or weakness this
could become a problem, and your question a point to consider.
Janis
> > Just curious, but why Valks and Barbs, rather than Valks and
> > Samurai? All three are powerful in melee, Valks and Samurai
> > have the advantage of being able to effectively use a ranged
> > weapon as well, and early speed, and the Samurai starts with
> > pretty good armor as well. About the only disadvantage of a
> > Samurai is that its quest nemesis has the Tsurugi of
> > Muramasa, which can result in an instant death if you aren't
> > careful.
[...]
> For the first question, I think both Valkyries and Barbarians
> have small advantages over Samurai in the early game. Barbs
> get their poison resistance, Valks get easy Excalibur,
Valks only get easy Excalibur if they're lawful, which isn't
guaranteed. While it is for Samurai. (Obviously, a Samurai has
to find a long sword first, but I've never found this to be a
problem.)
Pick "dwarf" at race screen. Instant lawful Valk.
> While it is for Samurai. (Obviously, a Samurai has
> to find a long sword first, but I've never found this to be a
> problem.)
Barrow wights always carry long swords, for reference.
The assumed-player under discussion is spoiled, so in this case it would
know about Ashikaga and the chance of instadeath.
Yes.
And thus preparing to kill him, e.g. by blessed paralysis and
free action. ("[...] you'll have all the time you like to get
what you need before you decide to meet them.")
Other nemeses would allow for more and therefore even simpler
to obtain options.
Janis
Or that common sense is wrong. Common sense says wizards are easy because
for much of the game they are so very powerful, but that neglects to
consider that "much of the game" does not include the opening game where
most danger lurks. It's there that a wizard's initial equipment crapshoot
exerts most influence - and a crapshoot is not a recipe for near-100%
ascension rates. The Valkyrie never gets very lucky with starting
equipment (not that getting a lamp isn't nice) but she never gets very
unlucky either.
--
David Damerell <dame...@chiark.greenend.org.uk> Distortion Field!
Today is First Brieday, March.
How are we defining "easiest"? Do we mean which class has the best
chance of a 100% ascension rate? Do we mean which class has the
easiest time getting through the game, even if they don't have the
highest potential ascension rate?
For example, if I was going for highest ascension rate, then I might
go for Orcish Wizard, just because they start with poison resistance
and magic resistance. They aren't the easiest choice to get through
the game, but they can survive some of the rare early insta-deaths
that other characters cannot survive. I would spam the crud out of
Elbereth, and play with such a high paranoia that the game would be
painfully slow and boring.
If you are defining easy as "least amount of though required", then
I'd pick a dwarven valk. Excalibur + good armor + lots of HP = easy
mode. valks can get a decent ascension rate without much thought
required. I wouldn't play with a high level of paranoia, so my
ascension rate wouldn't be as high, but the games I would ascend would
be really easy to do.