馃攳
Why Wideband Delphi Works (Agile Estimating) - #WhyShaiWhy Ep 53 - YouTube
Channel: unknown
[0]
Morning warriors, welcome to #WhyShaiWhy
[2]
For this week today, I want to introduce
[4]
to you a video that we got built for
[7]
planning poker, basically. And actually
[11]
you know what, what we'll do is I'll show
[13]
you the video first and then we'll go
[14]
back into the theory as to why planning
[16]
poker works because at the moment it's a
[18]
bit of a hot topic in agile, where in our
[21]
community, there are certain circles in
[23]
our community that say, Hey there should
[25]
not be any estimates for planning poker
[27]
sessions at all. Some are saying that you
[29]
should not use story points versus
[32]
hours versus story points or
[34]
t-shirt sizes. So we got to
[38]
thinking about this. We built a video
[39]
around this so let's have a look at that
[41]
first and we'll come back to it
[47]
Hey there! I see you got yourself the
[51]
Agile-as-a-Service Planning Poker deck.
[54]
Good choice!
[55]
Let me show you how to use
[57]
it!
[57]
Start off by choosing an estimation
[59]
method. With our Planning Poker deck, you
[62]
can estimate through Story Points, Hours
[64]
or T-shirt sizes. Once that you have
[67]
agreed on an estimation method, each team
[69]
member now takes a deck of cards. One
[71]
user story is then picked to be read out
[73]
and reviewed. Discuss technical and
[76]
business aspects of the story and
[78]
outline any risk or similarities with
[80]
previous work. Now on the count of three
[83]
everyone presents the card with the
[86]
number they think this user story should
[88]
be estimated at. Three, two, one - Go!
[91]
Everybody reveal their chosen card!
[95]
Most likely you have different estimates.
[97]
This is normal as different people have
[99]
different strengths and experiences.
[102]
It is really important not to take an
[104]
average of the team and set that as the
[106]
final estimate. If you have differences
[109]
discuss why some people are higher and
[111]
others are lower. Get a shared
[113]
understanding of some of the things each
[115]
individual thought about then do the
[117]
3-2-1-Go process again. When you
[120]
have a final estimation for a user story,
[122]
go on to the next story! Oh and here are
[126]
a couple of things to keep in mind.
[128]
Don't spend too much time on any one story.
[130]
These are just estimates and should be
[132]
treated as such.
[133]
Try to do user
[134]
stories in the same epic or theme
[137]
together so everyone has a shared
[139]
understanding in context.
[142]
Try to invite product owners, business analysts or even
[144]
business stakeholders to the session to
[146]
answer any questions the team might have.
[148]
Run planning poker regularly so you
[151]
don't have to estimate too many stories
[153]
at once.
[154]
Estimation fatigue is a real thing!
[156]
If you have any questions about
[158]
planning poker or agile in general, you
[161]
can contact us through e-mail or phone
[163]
us at 1 300 585 516
[166]
Softsolutions
[168]
Agile-as-a-Service
[169]
Okay so now you know how it
[172]
works, let's dive into why it works,
[175]
basically. Ok so this Delphi method,
[178]
this thing was built by the RAND
[179]
competi---
[181]
the RAND Corporation, which was basically
[185]
a think tank, a policy think tank
[186]
specifically, thinking about military in
[191]
this case during the Cold War.
[194]
And these were built in the United States.
[196]
So you can think about the way this
[199]
was playing out, right? There was the
[201]
1960s they were trying to figure out,
[204]
they were about to get into this arms
[206]
race, right? With an escalation of
[207]
technology, also known as weaponry, was
[212]
going to take force and it
[214]
was going to be a US vs Russia type,
[216]
bilateral domination in technology and
[222]
really, what the US or RAND
[225]
Corporation wanted to know was - what do
[228]
we need to do to win this race? This arms
[231]
race? And so basically what they did was
[234]
they fell down onto this idea of the
[236]
Delphi method. Now, the Delphi method
[240]
you guys saw in the video is
[241]
basically, you get a panel of experts in
[244]
the room. Let's say a bunch of generals,
[246]
in this case. And get them to estimate
[248]
something. So let's say in this case it's
[250]
the amount of hardware required to
[252]
overcome an enemy in any given conflict.
[255]
So the method was shown to be
[258]
quite close to the approximation,
[260]
the estimation was very close to the
[263]
actual. And therefore it started getting
[266]
a lot of popularity. In other things as
[271]
well, as time went by. So people
[273]
started forecasting things, like the
[276]
technology around organ transplantations,
[278]
oral contraceptives
[282]
self-replicating molecules, desalination
[286]
of water - like all these things we're now
[288]
getting forecasted using the Delphi
[290]
method or the Delphi method. And
[292]
what ended up happening, there was a 2008
[296]
review, i think, that look back at these
[299]
Delphi method estimations and
[301]
what we found was, they were actually
[302]
quite close to the approximations that
[305]
they had made back then. On
[309]
the flip side of course there were some
[310]
things that just blew out, even now the
[313]
Delphi method
[315]
out of the water, one was the the ability
[318]
of getting men on Mars. I think we've
[321]
missed that by a bit. Although that started
[324]
to pop a little bit. Other one was the world
[326]
population, I think that escalated a lot
[328]
higher than we expected it to. Anyway
[331]
that's the Delphi method. What we started
[335]
using for planning poker is called
[337]
the Wideband Delphi method and we added
[340]
the word Wideband because basically
[342]
what we're doing is, we're denoting the
[345]
high level or the high bandwidth of
[348]
communication required to get that
[351]
estimate out. So like you saw in the
[353]
video, we ask our panel of experts to
[356]
discuss the differences between the
[359]
estimations on their cards rather than
[362]
just sort of go with an average for
[363]
example or go with the mode which is the
[365]
most occuring card. You don't want that
[368]
sort of thing because what you're
[369]
introducing instead is communication for
[372]
people to understand, create a shared
[374]
understanding and then re-estimate to a
[377]
point where everyone's comfortable with
[378]
that estimate. That communication is why we
[382]
pull it like that. Okay so guys that's
[385]
the history of planning poker and if you
[387]
are in the No Estimates camp or in the
[389]
Story Points camp or in the Hours camp
[391]
or even the T-shirt Sizes camp, awesome.
[394]
But I did want to go back and start
[397]
thinking about why do we do this and
[399]
how do we do this. If you decide to go down
[404]
the estimation path guys, please feel
[406]
free to come and hit us up for a
[409]
Planning Poker deck of cards.
[413]
These, we will ship to you and your
[416]
team. (I was gonna say family) If your family wants
[419]
some, fine, no worries. These have what you
[423]
need and also on the back you'll see the
[424]
links to our video as well. So if you
[426]
would like to start thinking about
[428]
estimations, please get in touch.
[430]
Till then guys! Until next week! Cheers!
[432]
Have a great week!
[434]
you
Most Recent Videos:
You can go back to the homepage right here: Homepage